Hi Tom:
I think your observations are very astute. In the continually ticky/manual deticking example I mentioned earlier, what I noticed was the reduction of high frequency steady-state surface noise after I deticked the side. Since I was only affecting the ticks, that shouldn't have happened if the algorithm simply replaced the tick with synthesized clean audio from either side of the tick, which is what I've always been told the software did.
I completely agree that the most important factor is using the best existing disc source. I've written this before this before, but as a metaphor, think of it as the difference between a vaccine and a cure. Far better to get a vaccine and not get sick, than get sick and need to be cured.
--Jon Samuels
--- On Thu, 6/20/13, Tom Fine <[log in to unmask]> wrote:
From: Tom Fine <[log in to unmask]>
Subject: Re: [ARSCLIST] remastering discs using CEDAR, et al
To: [log in to unmask]
Date: Thursday, June 20, 2013, 8:59 AM
Hi Jon:
I've been surprised with the de-tick tool in Sony Soundforge. I know what you mean about removing sparkle from acetates and shellacs. I think many de-ticking tools incorporate some kind of surface-noise removal, too, the idea being that then the tick removal won't pump the background noise up and down. What I found with the Sony tool is that if you tweak the amount of tick level-reduction, you find a point where the ticks are minimalized enough not to be annoying and there is no pumping on the surface noise. I generally turn off the surface noise reduction part of the tool. I prefer to maximize music and minimize disk noise in the analog domain by careful selection of media (ie find the best copy available), good cleaning process, proper stylus and then tuning turnover and rolloff to ear. I don't really put too much stock in strict adherence to allegedly-used recording curves of yore. If they were actually used, then playback will obviously sound best in
that curve. Most of the time, either the equipment was not in perfect adjustment or an altogether different curve was used. Given the frequency limitations of much early disk-recording equipment, especially field equipment, I listen carefully to where the music actually is and try to minimize the noise in the surrounding frequencies. It's obviously hardest with well-recorded orchestral music and easiest with midrange-heavy content like a blues guy singing with his own guitar. With spoken-word material, I tune to audibility of the words above all else. I don't really care if it sounds like a phone line as long as the words are crystal clear, but it's rarely necessary to go that far!
-- Tom Fine
----- Original Message ----- From: "Jon Samuels" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Thursday, June 20, 2013 8:41 AM
Subject: Re: [ARSCLIST] remastering discs using CEDAR, et al
In my opinion, using deticking algorithms is a mixed bag. I've used CEDAR, Izotope, NoNoise and other declicking and dekrackleing software, and have almost never found there to be no negative sonic effect. No engineer worth his/her salt will ever get to the point of introducing digital artifacts, but the sound becomes homogenized and loses its' sparkle long before that. Some engineers feel that you can EQ to compensate for this loss. I think that's unadulterated nonsense.
Still most consumers feel that the trade-off is worth it, and there are time when the surface ticks are so extreme that no alternative is possible.
If one has the time (and energy), a superior way to detick is manually, one tick at a time. This is a nightmare to do on a ticky acetate, but the results are sonically better. Interestingly, I did a test of an continually ticky 78 acetate, and discovered that even manually detick algorithms affect the sound.
--Jon Samuels
--- On Thu, 6/20/13, Tom Fine <[log in to unmask]> wrote:
From: Tom Fine <[log in to unmask]>
Subject: Re: [ARSCLIST] Benny Goodman Carnegie Hall 1938
To: [log in to unmask]
Date: Thursday, June 20, 2013, 5:19 AM
One thing to remember is that current-vintage CEDAR and other restoration software is a completely different animal from the early-CD-era stuff over-used and, unfortunately, boasted about. A good comparison is the current Mosaic set of Duke Ellington small-group sessions vs the Sony "Duke's Men" CD sets. Pick any tune and hear clear differences between 1980s methods and modern methods. The latest Robert Johnson box set vs the massive-selling early-90's set is another example. The software and methods today, when used properly (alas, some engineers -- including guys who get a lot of work in this field -- still over-use the tools to the point of producing unpleasant artifacts), do a much better job of removing unwanted noises and keeping more music. It still comes down to hard work on the part of the engineer -- selecting the best playback method, finding the best media to playback, and using his ears and taste at every step.
There was just a good panel on this topic put on by AES NYC Section. Seth Winner had a particularly good demo, including videos showing how he cleans disks and finds just the right playback stylus before running one second of digital capture. Doug Pomeroy described in detail modern disk-transfer methods vs what he experienced at Columbia in the 70s (suffice to say, methods and technology have vastly improved). And Andreas Meyer touched on disk and tape playback plus played some before and after work he did on RCA opera reissues. I thought it was one of the best AES NYC meetings in years. People stayed around talking until we literally got kicked out of the room.
My final word on all of the modern tools -- and this is one man's opinions -- is that they still can't be automatic-thrown into a situation unless one is OK with many annoying digital artifacts. Each disk side, each reel of tape, must be considered in and of itself. The analog transfer quality is much more important than the arsenal of digital tools. And the engineer must listen very carefully for unwanted artifacts. To my ears, and perhaps others', digital artifacts are much more annoying than analog noise or ticks and pops. The two analog noises that I find equally annoying to digital artifacts are fuzz-distortion (which can actually be mitigated, to a point, with modern digital tools), crackle (which can be removed to a surprising extent in many cases) and hum (which has always been removable since the first notch filter was invented).
-- Tom Fine
----- Original Message ----- From: "Paul Stamler" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Thursday, June 20, 2013 2:55 AM
Subject: Re: [ARSCLIST] Benny Goodman Carnegie Hall 1938
> On 6/19/2013 7:10 PM, Clark Johnsen wrote:
>> "CEDAR... to restore the warmer sound of the original LPs." Oh dear. I
>> don't like the sound of that (as it were).
>
>
> Well, if not abused, Cedar is usually less intrusive than other cleaning-up software. I personally prefer careful descratching, no decrackling, careful filtering of infrasonic garbage, and no further processing. But that's just me.
>
> Peace,
> Paul
>>
>> On Wed, Jun 19, 2013 at 5:55 PM, Donald Clarke <[log in to unmask]> wrote:
>>
>>> Don't know anything about the Phoenix, but the Jasmine set is quite
>>> wonderful; I doubt if the quality of that transfer will be exceeded. A
>>> British company, transfers done by a Swede (Bjorn Almstedt), CDs made in
>>> Czech Republic; the notes say that after 18 months and close to 200 studio
>>> hours of work using "CEDAR and countless hard disc edits to restore the
>>> warmer sound of the original LPs without the clicks" is the result, but I
>>> had the vinyl set in the mid-1950s and this is superior to what I remember,
>>> plus it's more complete. Some said it was still not complete, a bit of
>>> somebody's solo missing, but I can't remember what it was. Where is Phoenix
>>> Jazz? I would be suspicious.
>>>
>>> Donald Clarke
>>>
>>>
>>> On Jun 19, 2013, at 5:17 PM, Thomas Stern wrote:
>>>
>>> I notice a forthcoming "Phoenix Jazz" 2-CD set, billed as COMPLETE
>>> recording.
>>> Sony, Jasmine and some other labels have also issued these recordings.
>>> Anyone have a
>>> detailed breakdown of differences if any, sound quality of transfers,
>>> editing, annotation, etc.
>>> of the various releases ???
>>> Thanks!
>>> Best wishes, Thomas.
>>>
>>
>
|