to DSP or not to DSP (1 Viewer)

[JohnRice]

All I can figure is we're mixing up the raw audio data rate and total data rate including things like subcode, error correction, etc.

 

[Manuel Delaflor]

"2. The other sees a recording as a starting point to play with until you get something that sounds "good" to you, regardless of what it sounded like on the studio monitors. "

Actually, I think is more than that. The goal is to restore the data which become lost in the recording process. Agreed, something which just "sound better" to someone is not necessarily good for other persons, but I feel that something that makes a recording to sound more natural is welcomed.

Of course, mileage can vary and not all DSP solutions will produce good results. One additional complication is that every recording will sound different, so a setting for one recording can do horrible things with other.

DSP is in its infancy, but I surely think that in the near future it will become more and more a standard than a curiosity.

 

[John Royster]

Manuel,
Recreating the recording to its natural state is by definition NO DSP and TWO speakers. We have two ears, the instruments are mixed and recorded for play with two speakers.
There is no right or wrong answer here, but to make statements like "more natural" and "restore what was lost" by adding/subtracting from the recording is contradictory.
I may be wrong here but I thought concert recordings and what not are about as pure as you can get. Two microphones are placed in one of the "good seats" in the house. So the microphones are truly recording the sound that would reach your ears complete with room accoustics, reverb and everthing else.
-edit- didn't mean to come off harsh but i'm one of those guys in camp #1. and pretty passionate about music/this darned hobby as well.

 

[David Judah]

The goal is to restore the data which become lost in the recording process

That is what DVD-A & SACD are for, not DSPs--no brickwall filter at 22kHz and more of the quality that was originally captured in the studio(provided they used the right equipment).

In a Home system, juicing up Red Book or other limited formats with DSP's is one option, but it can't hold a candle to the quality that is already inherent with the two high-rez formats on the market. If one wants to use DSPs to spice up a CD that was poorly recorded then I say go for it(although make sure that speaker placement and/or the room isn't the problem).

For computer sound the hardware and the software are most often both limiting factors, so using DSPs to spice things up can be beneficial, IMO, but it's a seperate issue from a Home Theater/Music system.

Ric P.,

Isn't your use of the TDS202 similiar to what Manuel is trying to acheive by using DSPs--that is, to add some ambience to the recordings?

DJ

 

[RicP]

Isn't your use of the TDS202 similiar to what Manuel is trying to acheive by using DSPs--that is, to add some ambience to the recordings?

Good question. But it's not really the same thing. The TDS202 is an analog device firstly, it applies a simple harmonic effect to an analog signal. In fact the TDS202 is much more akin to using a tubed pre-amp than any sort of digital manipulation of the incoming signal.
I've yet to see or hear any DSP mode that doesn't affect the signal for the worse (excluding of course things like 5 and 7 channel stereo, which really aren't DSP modes in the true sense of the term). There's no way that a piece of silicon can know what the original performance included, so therefore there is no way for it to restore anything. All it can do is apply a pre-determined digtial algorithm to a digital sample in a hope that it will apply a pleasing sound. More often than not it fails, and fails badly.
Tubes and the TDS202 on the other hand are adding harmonic overtones to an analog signal, which is something that is usually lost in the transformation to digital. The issue is that you cannot simulate what tubes or the TDS202 do digitally, they are inherently an analog process.

 

[RicP]

All I can figure is we're mixing up the raw audio data rate and total data rate including things like subcode, error correction, etc

Not really. I misread your 150KBs to be 150kbps because we typically do not use the KB identifier. When data rates get up over 1Mbps it is customary to describe it in terms of Mbps.
Anyway, the PCM audio data rate of a CD is 176.4KBps or 1.4112Mbps
So, now I see what you were saying. You were saying that a 24/96 digital master, run through some sort of "psychoacoustic" processing to reduce the data rate to ~ 1.4112Mbps would sound good. You know what? You're absolutely right. You know what else? This is virtually exactly how every CD is produced today.
An analog master is captured at 24/96 or 24/192, or an existing digital master of that rate is taken, and then it is run through one of several systems. Sony has SBM (Super Bit Mapping) which reduces a 20 or 24 bit PCM master (or even a DSD master) to the 16 bits needed for PCM. There is also the db Technologies Model 3000 dithering processor that accomplishes the same thing.
When they make a 16/44 CD from a 24/96 master, they don't just toss the extra bits and samples, they use a dithering station to reduce the data rate while retaining the most significant data to the waveform. This too, is a form of "psychoacoustic" processing and is how almost every CD is produced today.

 

[Manuel Delaflor]

David:

"So the microphones are truly recording the sound that would reach your ears complete with room accoustics, reverb and everthing else."

True, but I think it is also true that the mics are not picking exactly what we hear. The reason of this is that the receptor of the sound waves plays a role in the way they actually sound.

The mics transform air vibrations on electrical impulses, then this impulses are transformed in yet another domain. All this color the sound.

Our ears also convert the vibrations on electrochemical impulses, but then we have a brain that also "colors" the original waves in a much complex way a recorder can do.

Adding some reverberation, delays and other more complex effects (like harmonics restoration) to a two channel recording can add some of the information that was lost in the recording process.

_____________________________________________

RicP:

"I've yet to see or hear any DSP mode that doesn't affect the signal for the worse (excluding of course things like 5 and 7 channel stereo, which really aren't DSP modes in the true sense of the term). There's no way that a piece of silicon can know what the original performance included, so therefore there is no way for it to restore anything. All it can do is apply a pre-determined digtial algorithm to a digital sample in a hope that it will apply a pleasing sound. More often than not it fails, and fails badly."

Yes, it is really difficult to achieve a pleasant result, not to tell about the impossibility to use a general setting as an all around solution for every recording.

And about the information, there is no need to know how the sounds sounded before the recording was made, you can still enter data which will be a complement of the original recording. True, it will be not the actual information that was in the recording location, but a fairly good resemblance of it.

 

[Manuel Delaflor]

"That is what DVD-A & SACD are for, not DSPs--no brickwall filter at 22kHz and more of the quality that was originally captured in the studio(provided they used the right equipment). "

You are right about the 22Khz limitation, however, there is something called "Artificial Harmonics" invented in 1955 by a guy called Charles D. Lindridge. Theoretically, you can restore the information above the 22Khz range with this technology which is part of the arsenal of tools in the DSP arena.

 

[RicP]

Theoretically, you can restore the information above the 22Khz range with this technology which is part of the arsenal of tools in the DSP arena.

Theoretically being the key term there. It is extraordinarily difficult to simulate analog harmonics in a digital algorithm. I've never heard it done properly...never.

Perhaps you can tell us exactly which DSP modes you are referring to Manuel? Which Brand of processor/receiver and which exact mode?

 

[Ryan Spaight]

Perhaps you can tell us exactly which DSP modes you are referring to Manuel? Which Brand of processor/receiver and which exact mode?

The original post was referring to Winamp (Windows MP3 player) plug-ins.

Ryan

 

[Philip Hamm]

It also referred to a $25 external analog signal processing device called "The WOW thing" which adds a bass boost and an effect that sounds like "Ambience" on an old 80s Panasonic Boom Box. Look for the link on page 1 in one of my posts. Interesting little box.

 

[Manuel Delaflor]

Yes, I started by using those SRS products. But then I started to look at the other DSP plugins available for Winamp and I discovered that a friend uses a Behringer Dualfex Pro EX2200.

He lives at about 160 miles from where I live, so it will take sometime before I listen to it, but I think it surely deserves an evaluation.

He is extremely happy with it, but still he just use it with some recordings, mostly Jazz, and not with all his collection. In his words "it can add more acoustic space between the instruments", and this guy have relatively good equipment, Bryston separates driving NHT 3.3 (don't remember the actual model, but it was the best NHT had some years ago).

I just ordered the little Wow thing. It is only a toy, but I hope to learn if DSP technologies are for me or not, when hearing my CD's collection on my Eico tube amplifier/Klipsch Heresy/Sunfire Junior gear.

 

[Manuel Delaflor]

"Theoretically being the key term there. It is extraordinarily difficult to simulate analog harmonics in a digital algorithm. I've never heard it done properly...never."

I agree completely, DSP and Psychoacoustics are in their infancy, but looks promising, so far. Of course, with the arrival of SACD and DVD audio it is not very clear what the future of hi-quality audio will be.