IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
Comparing 78rpm transfers, what to do when the original 78 is not available
verdemar
post Feb 9 2010, 22:52
Post #1





Group: Members
Posts: 42
Joined: 10-February 06
Member No.: 27674



Hi, I am about to do some comparisons of various CD releases of music transferred from old shellacs. Before starting out, I could need some input on what to listen for and suggestions about objective measures of sound quality. I don't have the original 78s myself, which makes it a bit of a mess. On some cds the sound is filtered and eq'd a lot, on others not.

What I have noticed after following discussions here is that level matching is important, as it is easy to be fooled by a loud release, are there other things I should make sure to adjust before starting out?

If we assume tinkering a little with the sound is ok, applying eq, changing pitch/speed etc, what are good approaches to finding the most correct settings?
-I know some transfers are done at different speeds, is there some way to reliably find the correct speed/length of a song?
-If the wrong eq curve is applied to the music, it should theoretically be easy to fix it, but the trouble is I don't know how the music was processed.

And collecting the 78s doing the transfers myself is not really an option, I'm afraid.

Grateful for any help!
Go to the top of the page
+Quote Post
AndyH-ha
post Feb 10 2010, 07:58
Post #2





Group: Members
Posts: 2224
Joined: 31-August 05
Member No.: 24222



You are going to “do some comparisons of various CD releases of music transferred from old shellacs.” It seems to me you have to have well defined goal in order to make a plan for getting there.

For instance, without outside influences, my goal, if “making comparisons” would be to decide which I like and which I don’t like. It is conceivable that two, or more, versions that are easily different from each other might be equally enjoyable. Would I care if any came closest to attending the original performance? Possibly somewhat from intellectual curiosity, but in the circumstances you describe, I see no way to answer that. Therefore I would not fret over it at all.

If you can state a goal that is significantly different, then you have a more reasonable basis from which to determine whether or not that goal is obtainable. You have to be able to make objective and quantifiable statements about what you want to accomplish before you have much chance of working towards it rather than off in some oblique direction. I’m not saying you don’t have such things in mind, but I don’t see them stated clearly in your post.
Go to the top of the page
+Quote Post
botface
post Feb 10 2010, 10:32
Post #3





Group: Members
Posts: 360
Joined: 14-January 08
Member No.: 50483



QUOTE (verdemar @ Feb 9 2010, 21:52) *
Hi, I am about to do some comparisons of various CD releases of music transferred from old shellacs. Before starting out, I could need some input on what to listen for and suggestions about objective measures of sound quality. I don't have the original 78s myself, which makes it a bit of a mess. On some cds the sound is filtered and eq'd a lot, on others not.


One problem facing anyone transferring 78s is the quality of the original. A good original won't need much doing to it to make it sound presentable. A pressing of the same piece in poor condition will need a lot of processing and eq'ing to reduce the annoyance of the faults. If the the person producing the CD thinks the music is importat but only a poor quality original is available they may well just shrug and do their best but you will likely end up with an obviously "filtered and eq'd" sound

QUOTE
I know some transfers are done at different speeds, is there some way to reliably find the correct speed/length of a song?


This is very tricky and in many cases it will be impossible to tell. Your only hope of being sure is if there is a clearly audible instrument on the recoding which has a fixed key - EG many brass/wind instruments have Eb as their home key, I believe. Then, as long as you have some way of producing the right note you can check if the speed was correct when the transfer was done.

QUOTE
If the wrong eq curve is applied to the music, it should theoretically be easy to fix it, but the trouble is I don't know how the music was processed.


I understand that you want the eq to be right but I really wouldn't worry about it unless the result offends your ears in some way. Most of the collectors I know use "modern" magnetic cartridges and hi-fi systems to replay their 78s. Consequently they've been subjected to the standard RIAA eq used for LPs. But since that reduces the treble - and hence some of the noise - and boosts the bass a bit they actually prefer the result.

Go to the top of the page
+Quote Post
verdemar
post Feb 10 2010, 13:10
Post #4





Group: Members
Posts: 42
Joined: 10-February 06
Member No.: 27674



QUOTE (AndyH-ha @ Feb 10 2010, 07:58) *
You are going to “do some comparisons of various CD releases of music transferred from old shellacs.” It seems to me you have to have well defined goal in order to make a plan for getting there.


Thanks, my goal with comparing is to determine which is the 'best', which is not quite well defined, but finding the one that will take me closest to the original signal.

The reason I ask here is that I know I will be easily tricked by volume/eq and other processing, while another transfer might be better if corresponding treatment were applied. The problem is comparing an untreated and a heavily treated version made from different 78s.
Go to the top of the page
+Quote Post
verdemar
post Feb 10 2010, 13:20
Post #5





Group: Members
Posts: 42
Joined: 10-February 06
Member No.: 27674



QUOTE (botface @ Feb 10 2010, 10:32) *
One problem facing anyone transferring 78s is the quality of the original. A good original won't need much doing to it to make it sound presentable.


Good point, I think my problem maybe translate into trying to find which transfer was made from the best original. Unless it was filtered to death, that is... Tricky.

QUOTE
This is very tricky and in many cases it will be impossible to tell. Your only hope of being sure is if there is a clearly audible instrument on the recoding which has a fixed key - EG many brass/wind instruments have Eb as their home key, I believe. Then, as long as you have some way of producing the right note you can check if the speed was correct when the transfer was done.


Thanks, as far as I know, there wasn't really a clear standard for tuning of the instruments for the relevant material (tango music). The key instrument is the bandoneon, generally tuned to have A somewhere between 435 and 442.

QUOTE
I understand that you want the eq to be right but I really wouldn't worry about it unless the result offends your ears in some way. Most of the collectors I know use "modern" magnetic cartridges and hi-fi systems to replay their 78s. Consequently they've been subjected to the standard RIAA eq used for LPs. But since that reduces the treble - and hence some of the noise - and boosts the bass a bit they actually prefer the result.

Good point again, I tried it out a bit, and it seems to me it is not really that sensitive, but as you write I wanted to get it right if possible.
Go to the top of the page
+Quote Post
DVDdoug
post Feb 10 2010, 19:18
Post #6





Group: Members
Posts: 2711
Joined: 24-August 07
From: Silicon Valley
Member No.: 46454



I don't know that much about 78's, but I think you'll just have to use your ears and your judgment. I'd bet there are huge differences in standards and quality between the early recordings (around 1925) and the later recordings (around (1950).

Of course, the equipment was very limited and in the beginning there was no "high fidelity".

I wouldn't worry too much about playback speed. If it doesn't "sound wrong", Id assume it's correct. Even in the early 1900's they know how to make clocks and it wouldn't have been a big problem to get accurate recording/cutting speed. Players (gramophones) may not have been accurate, but I'd assume the recorders were. The only doubt I'd have, is that some very early records could have been made before the industry standardized the playback speed at 78 RPM... I don't know....

From what I've read, there were various EQ standards. I don't think you can resolve this one.... If you had the original, or if you can find-out what company published the original and in what year, maybe you could learn something. But, then you'll never know if the correct EQ was used by the person who did the digital transfer. Some records might have been EQ'd to play on an acoustic gramophone (the kind with a horn/megaphone "speaker" and no electronics). I'm not sure how you compensate for that! And, they were probably aiming for "clarity" and ""intelligibility" rather than fidelity & flat frequency response (much like the goals of the telephone system designers).

This post has been edited by DVDdoug: Feb 10 2010, 19:19
Go to the top of the page
+Quote Post
verdemar
post Feb 10 2010, 19:46
Post #7





Group: Members
Posts: 42
Joined: 10-February 06
Member No.: 27674



From the replies so far it seems it will all remain rather subjective... sad.gif

QUOTE (DVDdoug @ Feb 10 2010, 19:18) *
I wouldn't worry too much about playback speed. If it doesn't "sound wrong", Id assume it's correct. Even in the early 1900's they know how to make clocks and it wouldn't have been a big problem to get accurate recording/cutting speed. Players (gramophones)


The reason I brought this up is that there are transfers on the market with distinct speed differences (on the exact same recordings). Subjectively, I would say that mostly the slower sound the best, but it is hard to know if the people doing the transfers really knew what they were doing, as with the eq issue.

From what I have read elsewhere, it seems the recordings themselves were accurate, but the playback speed was not standardized. A few early records were made with a straight A played in the beginning for adjusting playback speed, but that is of course not on the cds, and I think it was a rare occurrence.

Thanks anyway.
Go to the top of the page
+Quote Post
AndyH-ha
post Feb 10 2010, 22:32
Post #8





Group: Members
Posts: 2224
Joined: 31-August 05
Member No.: 24222



There were actually quite a range of "standard" speeds, as well as a fair number of "standard" EQs. Every company had their own standards and these changed from time to time. There are published charts of many of these, but even with this in hand it can be difficult to tell because it isn't always easy to date the disk itself, to know where it belongs on the chart?

"Correct" playback demand a variable speed TT and a variable EQ preamp
http://www.vadlyd.dk/English/RIAA_and_78_RPM_preamp.html
http://www.phonostagepreamp.com/78rpm-riaa-equalization.htm
A handful of different sized 78 styli are necessary to find the best signal on the disk itself. The "best" transfer will generally depend more on the skill of the restorer than on the condition of the original -- peruse http://www.pristineclassical.com/
Go to the top of the page
+Quote Post
2Bdecided
post Feb 11 2010, 11:57
Post #9


ReplayGain developer


Group: Developer
Posts: 5359
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



People get the correct playback speed by matching the pitch of the music with the pitch they believe the original performers played at. You can't assume "78rpm" or "80rpm" (plenty of those!) means what it says - especially on early discs.

People get the "correct" EQ by starting with the markings on the disc where there's a known relationship between markings and EQ - see here:
http://home.clara.net/rfwilmut/repro78/repro.html#eq
...or by knowledge (e.g. this record company at this time has this equipment) or just by listening and guessing.

Even then, there's no guarantee that the rest of the recording chain had a flat frequency response (quite likely it didn't!), so people fiddle - and higher frequencies are usually removed to reduce audible background noise.


Lots of cheap re-issues (and some expensive ones) ruin the audio by:
1. bad declicking that shreds transients, percussive, brass and sometimes even vocal sounds
2. bad denoising that leaves the audio muffled, gated, and/or full of metallic artefacts
3. adding EQ, reverb, and even fake stereo effects that have nothing to do with the original recording


These latter ills are more damaging that slightly wrong frequency response or speed. You can easily fix the speed (and sometimes the frequency response) yourself to suit your own tastes.


FWIW most standard record players of the past played slightly too quickly. Everything I've ever owned from the 1950s-1980s without a speed control was fixed fast. All my records sound correct (but slow to my ears!) on a vari-speed deck correctly set with a strobe disc, or a decent deck that runs at the correct speed (e.g. LP12).


Spotify, Napster etc etc are great places to search for a popular vintage track and compare the highly variable transfer quality.

Cheers,
David.
Go to the top of the page
+Quote Post
verdemar
post Feb 11 2010, 20:10
Post #10





Group: Members
Posts: 42
Joined: 10-February 06
Member No.: 27674



QUOTE (AndyH-ha @ Feb 10 2010, 22:32) *
There were actually quite a range of "standard" speeds, as well as a fair number of "standard" EQs. Every company had their own standards and these changed from time to time.


That files under 'not standardized with me ;-)

Thanks for the information, I knew about the vad preamp, but not that they had so much relevant information on their website. The advice from there with respect to speed is to:

QUOTE (http://www.vadlyd.dk/English/RIAA_and_78_RPM_preamp.html)
gradually reduce the speed of a record until it starts to sound sluggish, and then increase it slightly (in my experience the ear is much more sensitive to low speeds than high speeds).
Go to the top of the page
+Quote Post
verdemar
post Feb 11 2010, 20:28
Post #11





Group: Members
Posts: 42
Joined: 10-February 06
Member No.: 27674



QUOTE (2Bdecided @ Feb 11 2010, 11:57) *
People get the correct playback speed by matching the pitch of the music with the pitch they believe the original performers played at. You can't assume "78rpm" or "80rpm" (plenty of those!) means what it says - especially on early discs.
(...)
...or by knowledge (e.g. this record company at this time has this equipment) or just by listening and guessing.


Generally I know both the recording date and the label, but not the key or pitch. Trouble is I don't know how the tranfers were made.

QUOTE
These latter ills are more damaging that slightly wrong frequency response or speed. You can easily fix the speed (and sometimes the frequency response) yourself to suit your own tastes.


Agreed, but the fixing will be subjective without a method to achieve 'correct' results. I wanted to see if anyone here had tips on how to do this more methodically, but it seems everybody basically just says 'use your ears' rolleyes.gif

QUOTE
Spotify, Napster etc etc are great places to search for a popular vintage track and compare the highly variable transfer quality.


I already have lots of material, including lots of duplicates, only differing by transfer, so I am trying to find a better way to choose which to use when playing for others, as they may not share my subjective judgements. But it seems there is no way around that.

For tango dance parties we usually group three or four songs together, and I usually choose from same disc or same series so that the sonic quality will be more or less the same within each set, correct or not. From what I've learnt so far that is probably the best approach. Also I wonder when it is worth the extra effort doing a little clean up on transfers with little or no declick/noise reduction versus using a transfer that has already been processed/filtered. The volume can be rather loud, so using 20s or early 30s output with no cleanup is not really an option, I think.


Go to the top of the page
+Quote Post
2Bdecided
post Feb 12 2010, 10:51
Post #12


ReplayGain developer


Group: Developer
Posts: 5359
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



I agree consistency is important in this context - including whether it's "restored" or not.

If you can run together several somewhat unrestored sides, it won't be as annoying as switching between restored and unrestored.

You've obviously got to do some restoration - running large clicks and pops through your system isn't good - general background noise might not be so bad if the crowd is already noisy!


Anyway, I'm intrigued - what kind of events are you doing? What kind of music are you playing? Any examples I can look up? DJing with 1930s music sounds cool!

Cheers,
David.

This post has been edited by 2Bdecided: Feb 12 2010, 11:09
Go to the top of the page
+Quote Post
verdemar
post Feb 12 2010, 18:17
Post #13





Group: Members
Posts: 42
Joined: 10-February 06
Member No.: 27674



QUOTE (2Bdecided @ Feb 12 2010, 10:51) *
Anyway, I'm intrigued - what kind of events are you doing? What kind of music are you playing? Any examples I can look up? DJing with 1930s music sounds cool!


Yes it's pretty cool. We are talking about tango music, the tango dance events are called 'milongas' and the music is almost entirely from the 30s, 40s and 50s with occasional songs from the 20s. The music is played in sets of three or four, with similar style of music and usually the sets ('tandas') are separated by intermission music ('cortinas'). So the cortina is an opportunity to clear the air before playing music with different sonic quality.

They have a little tango music at some of the streaming services, but not very much. A good reference for the classics of tango dance music can be found at http://www.tejastango.com/classic_tangos.html Make sure to check out the orchestras of Troilo, Di Sarli, Caló and Pugliese (mostly from 40s and onwards), for 30s: D'Arienzo, De Caro and Orquesta Tďpica Victor.
Go to the top of the page
+Quote Post
2Bdecided
post Feb 15 2010, 11:59
Post #14


ReplayGain developer


Group: Developer
Posts: 5359
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



Thanks verdemar. Found quite a few on Spotify - I hear the challenge you're having with sound quality! Transports you to a different era though.

Cheers,
David.
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 19th December 2014 - 04:20