Hello- For any mass restoration - which is what I am doing 7 days a week -
you want a process that is guaranteed to not introduce errors to the sound.
To achieve the noise reduction, you can run the decrackle - at least on
Dartpro - multiple times with the filter builder.
Decrackle does not affect the high frequencies as does the declick feature.
I don't audition the decrackle filter in this - I look for physical clicks
that will need to be removed and usually get under 5,000 interventions
before I start manually declicking
For the declick - what sounds like hiss on a 78 rpm record is commonly the
combination of many tiny clicks.
For this I have the filter builder run up to 3 times to deal with that kind
of noise. For this , I can deselect the filters in the filter builder - i.e.
for a record with very slight hiss, I might run it just once - I always
audition results for this and it is the last filter I run aside from the
dehiss.
Dehiss is where you can hear a steely quality if it is set too high - or in
the case of some records, it is not beneficial at all.
The less decisions you have to make in the process the better.
It is better to spend a little extra time this way and you won't get into
trouble and have to back up and reprocess with alternate settings.
Having higher settings might get rid of the noise faster, but when you get
to that really loud growly trumpet or a tenor like James Melton or Allan
Jones with a complex wave form, at the end of a record - you might have to
delete and back up with lower settings.
I don't feel that you can automatically know where the difficult areas of a
record are unless you're really familiar with the work. Just my opinion -
Mickey Clark
P.S. - Here are a couple of tracks by the Boston Symphony Orchestra
conducted by Serge Koussevitzsky recorded May 8th 1936 - issued on Red
Seal Victor 14230. These are both from the Damnation of Faust - Presto and
Waltz and the Rakoczy March by Berlioz
These were recorded with the processes mentioned
https://www.dropbox.com/s/jidcvvpgh7j2xf4/Boston%20Symphony.zip?dl=0
Mickey Clark
710 Westminster Avenue West
Penticton BC
Canada
250-462-7881
V2A 1K8
http://mcproductions.ca
1-250-462-7881
-----Original Message-----
From: Brewster Kahle
Sent: Saturday, June 30, 2018 9:26 AM
To: [log in to unmask]
Subject: [ARSCLIST] Mass restoration of thousands of 78rpm sides in the
Great 78 Project
You are all right-- we want to do the transfer super well in raw, but
we also really want to make pristine versions, which we have not done.
We are looking for methods to do mass restoration, what will scale to
hundreds of thousands. I am dreaming that the right engineers could
model the wear and back it out, and even model the distortions in the
recording and back that out, so we could even have truer sound than the
orginal (ok, a bit of a dream).
If anyone is interested in this, Jessica Thompson has been contracted by
the Internet Archive to investigate this (cc'ed).
-brewster
On 6/30/18 9:13 AM, Mint Records wrote:
> Because you don't archive that way. You preserve the sound of the original
> discs. Technology will move on and provide better and better restoration
> techniques. If you archive the restored version only you take away the
> chance of better results in the future. Can you also imagine the amount of
> work needed for 100000's of 78s? And of course it's double the storage
> space.
>
> As for "no sound quality is lost, only noise". That is open to question
> and
> is down to how sympatheticly its done. It really easy to damage the
> original sound. You can't rely on automatic settings.
>
> Archive.org are doing this to what has pretty much become the
> international
> standard for digitisation - unrestored transfer at 96k 24bit they also go
> the extra mile with different eq curves and styli.
>
> It's a great project.
>
>
> On Sat, 30 Jun 2018, 16:52 6295LARGE, <[log in to unmask]> wrote:
>
>> Great collection and effort, but why no click and crackle removal? No
>> sound
>> quality is lost, only noise.
>>
>> On Wed, Jun 27, 2018 at 1:37 AM, Brewster Kahle <[log in to unmask]>
>> wrote:
>>
>>> Paul-- thank you for posting the helpful article.
>>>
>>> Clark-- the condition of the 78's from the Boston Collection are quite
>>> good-- I am guessing they did not circulate, but they are not mint.
>>>
>>> We have a uncoming collection from a distributor, so they are mint.
>>> They are from the 40's and early 50's. So looking forward to those as
>>> well.
>>>
>>>
>>> We are looking for the best we can get.
>>>
>>> onward!
>>>
>>> -brewster
>>>
>>>
>>>
>>> On 6/26/18 4:25 PM, Paul T. Jackson wrote:
>>>> Here is a note about the Boston records as well as the 78s project
>>>> which include 20 other collections.
>>>> https://www.smithsonianmag.com/smart-news/boston-public-
>>> library-will-digitize-200000-vintage-recordings-180966931/
>>>>
>>>>
>>>> On 6/26/2018 3:33 PM, Clark Johnsen wrote:
>>>>> Just askin': Has any attention been paid to getting unworn or un-noisy
>>>>> sides for this laudable preservation effort? When I read just now
>>>>> that your
>>>>> source was a public library where patrons would have been borrowing
>> them
>>>>> and/or playing them on any sort of machine, it made me wonder. By the
>>>>> way,
>>>>> are you in Boston?
>>>>>
>>>>> On Tue, Jun 26, 2018 at 3:52 PM, Brewster Kahle <[log in to unmask]
>>>>> wrote:
>>>>>
>>>>>> for the fanatics (well, like me), we have another 6,000 78rpm sides
>> up
>>>>>> on archive.org as part of the Great 78 Project. (now 74k digitized
>>>>>> <https://archive.org/details/georgeblood> total)
>>>>>>
>>>>>> I was able to find dates for 91% of them, and volunteers have
>>>>>> started on
>>>>>> others.
>>>>>>
>>>>>> Lots of Victors and Deccas, almost all from the Boston Public Library
>>>>>>
>>>>>> link to the new ones:
>>>>>>
>>>>>>
>>>>>> https://archive.org/details/georgeblood?and%5B%5D=
>>>>>> publicdate%3A%5B2018-06-10+TO+*%5D&sin=&sort=-publicdate
>>>>>>
>>>>>> onward!
>>>>>>
>>>>>> -brewster
>>>>>>
|