Noise reduction

[Alfred001]

I'm working on a paper for school about digitization of sound recordings on analogue discs, and I'm completely ignorant on these matters so could you lend me a hand?

I'm using this source:

http://www.bl.uk/reshelp/findhelprestyp ... ration.pdf

It is quite convenient in it's comprehensibility, but judging from some of my other smaller sources it is quite out of date (the gentleman who was writing it passed in 2006 before finishing it).

Does anyone know a more up to date source (or sources) I could use? It doesn't have to be too detailed and comprehensive, but something to give me an idea.

[Alfred001]

EDIT: I had a specific question about reducing hiss (I mislabeled the thread), I forgot to write it. In my source it says:

The hiss is always the ultimate limit beyond which we cannot go on a single copy, so we must make every effort to eliminate it at source. In fact, steady hiss is not usually the most noticeable problem, but rather crackles and pops; but we shall see later that there are ways of tackling those. It is the basic hiss that forms the boundary to what is possible.

If I'm reading this right the implication is that there isn't (or wasn't at the time) a way to deal with the hiss once the music has been digitalized. I've read some things which suggest that this is no longer the case. Could someone tell me (or point me to a source) what the situation is currently?

[Trackside]

http://www.clickrepair.net/

[Hanuman]

I've only browsed the contents but it looks like a terrific and very useful paper. Thanks so much for linking to it. I don't think it's out of date at all - it's dated 2008 and, really, a great deal of the theory, and even practical application, has been understood for decades.

Regarding the question about noise, perhaps you are alluding to broad-band digital noise-reduction processing? This works by sampling a section of noise only and using that to create a noise profile. Reduction then takes place only at the frequencies in that profile and certain thresholds can be set to minimize the effect on actual program content. It's potentially really effective (I use Waves Restoration plugins to do this) but it is destructive and the "Holy Grail" of perfect noise reduction with no audible artifacts at all is elusive at best. I would say that, in practical use, the statement: "It is the basic hiss that forms the boundary to what is possible" is accurate.

Actually, I think the design purpose of this kind of processing is to remove stuff like air-conditioning noise from location film sound recordings but the technique is effective, as I say, in dealing with recording-media noise artifacts.

[Dimal]

Yes, agree...

That "Paper" is very interesting indeed. Thanks from me too...

Over the last 2-3 years, I've auditioned quite a few specialist noise reduction applications before finally settling on iZotope RX(2). It allows one a lot of flexibility in fine tuning the extraction of unwanted noise, while at the same time, minimising the addition of any unwanted artefacts. It also has the added benefit of being very easy to learn, and then use to great effectiveness.

It isn't the be-all, end-all of course and I doubt that any single application would ever be that. For example, its Click/Crackle/Pop removal algorithms leave a lot to be desired as I have been unable to reach a point where the offensive noise is removed, leaving the underlying music (audibly) untouched. For this work, I have ultimately settled on using "Click Repair" (linked above by "trackside") as it allows fine-tuning down to almost ridiculous levels of exactness, so that musical content is not affected. Takes a bit more of a learning curve to become proficient (not there yet ) but is well worth the trouble to do so, if one is serious about archiving one's precious vinyl collection.

For other, more basic digital editing, I use Audition 3.01 since I already owned that application after upgrading from "Cool Edit Pro" (which, as it turns out, wasn't really necessary, given how few of the "features" I actually use).

Anyway, hope some of this is helpful...

Mal.

[Alfred001]

Thanks for the help everyone!

[Hanuman]

EDIT: I had a specific question about reducing hiss (I mislabeled the thread), I forgot to write it. In my source it says:

The hiss is always the ultimate limit beyond which we cannot go on a single copy, so we must make every effort to eliminate it at source. In fact, steady hiss is not usually the most noticeable problem, but rather crackles and pops; but we shall see later that there are ways of tackling those. It is the basic hiss that forms the boundary to what is possible.

If I'm reading this right the implication is that there isn't (or wasn't at the time) a way to deal with the hiss once the music has been digitalized. I've read some things which suggest that this is no longer the case. Could someone tell me (or point me to a source) what the situation is currently?

I've been reading this "paper" (it's actually a 300 page book!) and now I'm far enough in to get a better understanding of the context of that quote.

It was made in specific reference to the steady hiss caused by the granular structure of records, mainly related to shellac 78s at that point in the book although pertinent to vinyl and other grooved media as well. The point seems to be that to maximize "power bandwidth product" (the importance of which was explained with emphasis right off the bat) the hiss must be minimized at source because, if I interpret the authors writing correctly, power bandwidth product only goes down through the processes, never back up. A number of strategies for minimizing steady groove hiss were described, related to parameters like the stylus profile and tracking force.

I must say that this read has been very, very educational (I'm not even a third of the way through it yet) and, frankly, I feel that every VE member who plans to make a post to the Vinyl Ripping forum (as well as a few others) should read this book as matter of duty. You've got the views of an esteemed BBC engineer and British Library Archivist on a number of thorny issues that are regularly debated here and that's got to be a great resource.

[Vinylmasters]

EDIT: I had a specific question about reducing hiss (I mislabeled the thread), I forgot to write it. In my source it says:

The hiss is always the ultimate limit beyond which we cannot go on a single copy, so we must make every effort to eliminate it at source. In fact, steady hiss is not usually the most noticeable problem, but rather crackles and pops; but we shall see later that there are ways of tackling those. It is the basic hiss that forms the boundary to what is possible.

If I'm reading this right the implication is that there isn't (or wasn't at the time) a way to deal with the hiss once the music has been digitalized. I've read some things which suggest that this is no longer the case. Could someone tell me (or point me to a source) what the situation is currently?

The best solution to removing his in the ORIGINAL recording I have found is simply transferring blank audio tape samples - with NO Dolby or DBX and such onon - and getting a sample of it, and then use that rather than the given sample. Most programs allow you to adjust the amount to be reduced as well - just use good headphones and you'll know how far you can go. It's a simple, yet easy, fix I've used numerous times.