|
|
Post by stormymondays on Apr 7, 2019 15:11:52 GMT -6
I have a daunting job ahead. I need to restore a couple dozen 78 RPM transfers. The transfers have already been done (they come from a library), so there is no way to optimize the analog capture. I need to deal strictly with declick, denoise, and EQ.
The discs were badly damaged, there's a lot more noise than signal. It doesn't help that they are field recordings of singers acappella, or backed with either a bagpipe or a distant piano.
I own the Sonnox suite, which is pretty good, but I think most people use Izotope RX. Any recommendations on that front? The advantage of the Sonnox denoiser is that it doesn't need to learn the noise pattern. RX does a LOT more but it looks like I would need RX 7 Advanced and that is over budget for this project. Anyway, I'm all ears on the software front, budget be damned!
My other question is if it would be better to EQ first for good sound and then declick and denoise, or the other way around? I'm pretty certain that removing clicks, pops and crackle goes before noise and hiss reduction, but not sure what's the right point of the process to EQ. Again, any suggestions are welcome!
|
|
ericn
Temp
Balance Engineer
Posts: 14,941 
|
Post by ericn on Apr 7, 2019 16:28:28 GMT -6
I have a daunting job ahead. I need to restore a couple dozen 78 RPM transfers. The transfers have already been done (they come from a library), so there is no way to optimize the analog capture. I need to deal strictly with declick, denoise, and EQ. The discs were badly damaged, there's a lot more noise than signal. It doesn't help that they are field recordings of singers acappella, or backed with either a bagpipe or a distant piano. I own the Sonnox suite, which is pretty good, but I think most people use Izotope RX. Any recommendations on that front? The advantage of the Sonnox denoiser is that it doesn't need to learn the noise pattern. RX does a LOT more but it looks like I would need RX 7 Advanced and that is over budget for this project. Anyway, I'm all ears on the software front, budget be damned! My other question is if it would be better to EQ first for good sound and then declick and denoise, or the other way around? I'm pretty certain that removing clicks, pops and crackle goes before noise and hiss reduction, but not sure what's the right point of the process to EQ. Again, any suggestions are welcome! In theory I would say EQ last but, in reality it’s what works. Do remember to make sure your cartridge has a 78 stylus and your working from a true mono signal. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 8, 2019 0:07:32 GMT -6
I have RX7 Advanced and am pretty sure it would do a great job on that. I'd say EQ last too. Eric, the transfers have already been done, so no need to worry about cart/stylii etc. Acon Acoustica is cheaper than RX Advanced, and supposed to be great, but not tried it myself. acondigital.com/products/acoustica-audio-editor/ |
|
|
|
Post by donr on Apr 8, 2019 1:52:03 GMT -6
I have a daunting job ahead. I need to restore a couple dozen 78 RPM transfers. The transfers have already been done (they come from a library), so there is no way to optimize the analog capture. I need to deal strictly with declick, denoise, and EQ. The discs were badly damaged, there's a lot more noise than signal. It doesn't help that they are field recordings of singers acappella, or backed with either a bagpipe or a distant piano. I own the Sonnox suite, which is pretty good, but I think most people use Izotope RX. Any recommendations on that front? The advantage of the Sonnox denoiser is that it doesn't need to learn the noise pattern. RX does a LOT more but it looks like I would need RX 7 Advanced and that is over budget for this project. Anyway, I'm all ears on the software front, budget be damned! My other question is if it would be better to EQ first for good sound and then declick and denoise, or the other way around? I'm pretty certain that removing clicks, pops and crackle goes before noise and hiss reduction, but not sure what's the right point of the process to EQ. Again, any suggestions are welcome! In theory I would say EQ last but, in reality it’s what works. Do remember to make sure your cartridge has a 78 stylus and your working from a true mono signal. I'm one of a few old timers here. I remember as a young kid, seeing and using 78 rpm record players, electric and acoustic, the latter already antique. Most had tone arms heavier than dinner knives, or tone arms jointed right at the end of the arm, instead of at the pivot point. And the metal stylus in the transducer was replaceable, with a thumb screw fastener. In fact, you'd buy a paper envelope full of replacement stylus at your record store, cause you'd need them. The stylus would wear out. Later, slightly lighter tone arms on multi-speed turntables had a cartridge with two styli, one for 78's and one for 331/3 and 45 RPM records. The stylus material went from all metal to sapphire or diamond tip. Flip it over for what you needed. The 78 records were harder and more brittle than the vinyl 45's and 33's of the mid 50's and beyond. I suppose the metal of the old time 'needle' was soft to reduce wear on the 78 record grooves. I don't know, I suppose the early record players were tough on the record pressing anyway. The tone arms of modern players became less and less massive, with precision down force and accepting standard mount cartridges from many makers. Play the records without wearing them out, and enhancing the transient response, and faithfully tracking the hi and low frequencies of the pressing with cartridge design. I was too young to realize it, that hi-fi was exciting in that time, because huge gains in analog fidelity and linearity were happening continually in real time in those years, right up until the digital revolution. |
|
ericn
Temp
Balance Engineer
Posts: 14,941
|
Post by ericn on Apr 8, 2019 1:54:31 GMT -6
I have RX7 Advanced and am pretty sure it would do a great job on that. I'd say EQ last too. Eric, the transfers have already been done, so no need to worry about cart/stylii etc. Acon Acoustica is cheaper than RX Advanced, and supposed to be great, but not tried it myself. acondigital.com/products/acoustica-audio-editor/I missed that when reading way to fast, but it is the stage where most 78 transfers are screwed up! Len Feldman, who did all the bench testing for Audio Magazine had this brilliant phono preamp that had all the different 78 EQ curves and a drift control for both the low and the high end so you could truely match the curve used to cut the master. For years I tried to convince plug-in developers that a digital version would be a great tool. |
|
Deleted
Deleted Member
Posts: 0
|
Post by Deleted on Apr 8, 2019 2:21:36 GMT -6
That would be a great tool Eric, someone should do it! AFAIK that is not part of any of the standard repair and restoration tools, when it definitely should be!
|
|
|
|
Post by stormymondays on Apr 8, 2019 6:13:54 GMT -6
Apparently these were captured with a Stanton turntable with a direct USB output. So: cheap converters, wrong stylus and most likely a RIAA curve that needs to be reversed. Suggestions?
|
|
|
|
Post by Ward on Apr 8, 2019 6:20:45 GMT -6
What about Absentia DX for Pro Tools?
|
|
|
|
Post by the other mark williams on Apr 11, 2019 0:32:26 GMT -6
Apparently these were captured with a Stanton turntable with a direct USB output. So: cheap converters, wrong stylus and most likely a RIAA curve that needs to be reversed. Suggestions? Yikes! How's the client? I hope they're not expecting a miracle. Or I hope their definition of "miracle" is obtainable for you. Still, sounds like it might be a fun project, actually! |
|
|
|
Post by stormymondays on Apr 11, 2019 1:13:04 GMT -6
Apparently these were captured with a Stanton turntable with a direct USB output. So: cheap converters, wrong stylus and most likely a RIAA curve that needs to be reversed. Suggestions? Yikes! How's the client? I hope they're not expecting a miracle. Or I hope their definition of "miracle" is obtainable for you. Still, sounds like it might be a fun project, actually! It IS a fun project! It's like solving a big puzzle. Fortunately the client is good - he's actually attempted to restore these tracks using Adobe Audition and came to a passable result, but I have already made his attempts bite the dust  I have all these moving pieces that could be placed in a different order, all with different results: -De-crackle -Undo RIAA curve applied on capture -Filter unneeded/non existent frequencies -De-noise -EQ to make it sound as good as possible The only thing that goes first for sure is the decrackle. I'm still not sure if I should EQ it to sound good and leave the final denoise for last, or EQ after denoising. Same for the filter. I can't compress because it makes the noise more prominent. |
|
|
|
Post by the other mark williams on Apr 11, 2019 6:25:18 GMT -6
Yikes! How's the client? I hope they're not expecting a miracle. Or I hope their definition of "miracle" is obtainable for you. Still, sounds like it might be a fun project, actually! It IS a fun project! It's like solving a big puzzle. Fortunately the client is good - he's actually attempted to restore these tracks using Adobe Audition and came to a passable result, but I have already made his attempts bite the dust I have all these moving pieces that could be placed in a different order, all with different results: -De-crackle -Undo RIAA curve applied on capture -Filter unneeded/non existent frequencies -De-noise -EQ to make it sound as good as possible The only thing that goes first for sure is the decrackle. I'm still not sure if I should EQ it to sound good and leave the final denoise for last, or EQ after denoising. Same for the filter. I can't compress because it makes the noise more prominent. i would denoise/decrackle first, mainly b/c once you EQ, you may be then accentuating some of the noisy frequencies with EQ. Generally it’s better to get rid of the noise first. Of course, it’s always worth trying both ways, b/c you never know... |
|
ericn
Temp
Balance Engineer
Posts: 14,941
|
Post by ericn on Apr 11, 2019 8:33:17 GMT -6
Yikes! How's the client? I hope they're not expecting a miracle. Or I hope their definition of "miracle" is obtainable for you. Still, sounds like it might be a fun project, actually! It IS a fun project! It's like solving a big puzzle. Fortunately the client is good - he's actually attempted to restore these tracks using Adobe Audition and came to a passable result, but I have already made his attempts bite the dust I have all these moving pieces that could be placed in a different order, all with different results: -De-crackle -Undo RIAA curve applied on capture -Filter unneeded/non existent frequencies -De-noise -EQ to make it sound as good as possible The only thing that goes first for sure is the decrackle. I'm still not sure if I should EQ it to sound good and leave the final denoise for last, or EQ after denoising. Same for the filter. I can't compress because it makes the noise more prominent. It really comes down to what works depending on track content and condition if you really want to get it right. I know of projects where different parts of a track were done with completely different processing to get them right. Years ago I did some transfers of a bunch of stuff done on one of those home record cutters and had to automate a bunch of stuff because I swear you could hear the cutting stylus dying as it cut through one side! |
|
|
|
Post by the other mark williams on Apr 11, 2019 10:52:07 GMT -6
It IS a fun project! It's like solving a big puzzle. Fortunately the client is good - he's actually attempted to restore these tracks using Adobe Audition and came to a passable result, but I have already made his attempts bite the dust I have all these moving pieces that could be placed in a different order, all with different results: -De-crackle -Undo RIAA curve applied on capture -Filter unneeded/non existent frequencies -De-noise -EQ to make it sound as good as possible The only thing that goes first for sure is the decrackle. I'm still not sure if I should EQ it to sound good and leave the final denoise for last, or EQ after denoising. Same for the filter. I can't compress because it makes the noise more prominent. It really comes down to what works depending on track content and condition if you really want to get it right. I know of projects where different parts of a track were done with completely different processing to get them right. Years ago I did some transfers of a bunch of stuff done on one of those home record cutters and had to automate a bunch of stuff because I swear you could hear the cutting stylus dying as it cut through one side! Wow. Super cool. I really love restoration and preservation work. |
|
|
|
Post by johneppstein on Apr 11, 2019 11:50:32 GMT -6
In theory I would say EQ last but, in reality it’s what works. Do remember to make sure your cartridge has a 78 stylus and your working from a true mono signal. I'm one of a few old timers here. I remember as a young kid, seeing and using 78 rpm record players, electric and acoustic, the latter already antique. Most had tone arms heavier than dinner knives, or tone arms jointed right at the end of the arm, instead of at the pivot point. And the metal stylus in the transducer was replaceable, with a thumb screw fastener. In fact, you'd buy a paper envelope full of replacement stylus at your record store, cause you'd need them. The stylus would wear out. Later, slightly lighter tone arms on multi-speed turntables had a cartridge with two styli, one for 78's and one for 331/3 and 45 RPM records. The stylus material went from all metal to sapphire or diamond tip. Flip it over for what you needed. The 78 records were harder and more brittle than the vinyl 45's and 33's of the mid 50's and beyond. I suppose the metal of the old time 'needle' was soft to reduce wear on the 78 record grooves. I don't know, I suppose the early record players were tough on the record pressing anyway. The tone arms of modern players became less and less massive, with precision down force and accepting standard mount cartridges from many makers. Play the records without wearing them out, and enhancing the transient response, and faithfully tracking the hi and low frequencies of the pressing with cartridge design. I was too young to realize it, that hi-fi was exciting in that time, because huge gains in analog fidelity and linearity were happening continually in real time in those years, right up until the digital revolution. <chuckle>
I remember all that, and one more - in addition to the little packets of replacement steel needles sold for the old 78 rpm players they also sold cactus needles - yeah, that's right, real cactus needles from one of the giant species such as the saguaro. The catus needles needed frequent replacement but didn't wear the fragile 78 rpm discs a badly as steel.
|
|
ericn
Temp
Balance Engineer
Posts: 14,941
|
Post by ericn on Apr 11, 2019 12:02:28 GMT -6
It really comes down to what works depending on track content and condition if you really want to get it right. I know of projects where different parts of a track were done with completely different processing to get them right. Years ago I did some transfers of a bunch of stuff done on one of those home record cutters and had to automate a bunch of stuff because I swear you could hear the cutting stylus dying as it cut through one side! Wow. Super cool. I really love restoration and preservation work. If you have the patience and skill restoration work can be very rewarding. There are good guys who have all these presets, the great guys just do what they have to. |
|
