On board sound card vs external sound card for vinyl capture

Quote:

Originally Posted by Pitou 

Hello,

 

My first post!

 

I just want to know if it's better to have an external sound card or just use my internal sound card in my PC for vinyl capture?

 

I did some test with a Roland Edirol FA-66 and was very satisfied with it. My onboard soundcard chip is an AD1988B which seems to be quite good.

 

My turntable is a Goldring and I listen to music with my Rotel receiver along with some B&W speakers. I also have a pair of Grado headphones.

 

So basically how does the AD1988B compare to the Edirol FA-66

 

Any opinions?

 

Thank you very much!

 

Pitou!

Depends on how it is implemented. I have seen in-circuit noise measurements on it (AD1988B) ranging from -62db to -87db or so, -87db is okay but nothing special (< 15 bits) , -62db is frankly awful. You can test your system with the righmark audio analyser (RMAA) . NB the incipient noise from an LP is relatively high and thus the extra noise added by even a mediocre ADC is unlikely to be noticeable. I'd stick with the Edirol

Quote:

Originally Posted by nick_charles 

 

NB the incipient noise from an LP is relatively high and thus the extra noise added by even a mediocre ADC is unlikely to be noticeable. I'd stick with the Edirol

This really depends on how the RIAA correction curve is implemented. 

If a dedicated phono-stage is used, and a line-level signal is fed into the soundcard, I agree the noise from the ADC is fairly insignificant here. 

If the cartridge is hooked up to a microphone input, and RIAA is done digitally noise in the ADC is VERY important. 

The OP didnt say how he is doing the RIAA correction, and I would assume he is using a dedicated phono-stage, but Its still worth repeating. 

Quote:

Originally Posted by Pitou 

I have to admit, I don't know what is the RIAA correction. Could you explain a bit?

 

RIAA correction is a very specific EQ applied to the signal BEFORE its put on the record. When you play it back the opposite EQ is used, and you get a flat response. 

It is not an EQ like someone might use an EQ to fix the fact that their crappy speakers have no bass. It adds X at the beginning of the process, and subtracts x at the end. The end result is exactly the same as what went in. 

Why we use RIAA:

Early in the development of the electronically-recorded record (and probably before), it was noted that the most annoying type of noise was pops and clicks. To solve this problem some smart dudes figured out that audibility of pops & clicks could be reduced if you EQ'ed the highs way high. It was also noted that if you had too much bass on the disc it caused weird problems - so these guys attenuated the bass before they cut the disc. When you play it back you attenuate the highs (which also attenuates the effect of dust - you like that?) and amplify the lows to get a flat result.  

Pretty much everyone agreed that this was a good idea in general, but nobody could agree on exactly how to implement it so at first there was a lot of confusion and disagreement between MFR's. This was kind of sucky because an RCA record might not sound right on a doiche-gramaphone-player.... Eventually RIAA came around and set up their standard that everyone agreed (or was forced) to follow so all records made after a certain time would play properly on anyone's gear.

So the problems with digital RIAA is that the signal directly out of the cartridge is ~40db higher at 20khz than at 20hz. This requires a boatload of manipulation in the digital domain which comes after the noise is introduced by the ADC which of course treats noise as signal and amplifies it right along. 

With an analog phono stage noise is still an issue (its always an issue - 95% of all problem solving in audio is simply moving the problem somewhere else) but the importance of the soundcard is reduced.

Quote:

Originally Posted by nikongod 

 

RIAA correction is a very specific EQ applied to the signal BEFORE its put on the record. When you play it back the opposite EQ is used, and you get a flat response. 

 

It is not an EQ like someone might use an EQ to fix the fact that their crappy speakers have no bass. It adds X at the beginning of the process, and subtracts x at the end. The end result is exactly the same as what went in. 

 

Why we use RIAA:

Early in the development of the electronically-recorded record (and probably before), it was noted that the most annoying type of noise was pops and clicks. To solve this problem some smart dudes figured out that audibility of pops & clicks could be reduced if you EQ'ed the highs way high. It was also noted that if you had too much bass on the disc it caused weird problems - so these guys attenuated the bass before they cut the disc. When you play it back you attenuate the highs (which also attenuates the effect of dust - you like that?) and amplify the lows to get a flat result.  

 

Pretty much everyone agreed that this was a good idea in general, but nobody could agree on exactly how to implement it so at first there was a lot of confusion and disagreement between MFR's. This was kind of sucky because an RCA record might not sound right on a doiche-gramaphone-player.... Eventually RIAA came around and set up their standard that everyone agreed (or was forced) to follow so all records made after a certain time would play properly on anyone's gear.

 

 

So the problems with digital RIAA is that the signal directly out of the cartridge is ~40db higher at 20khz than at 20hz. This requires a boatload of manipulation in the digital domain which comes after the noise is introduced by the ADC which of course treats noise as signal and amplifies it right along. 

 

With an analog phono stage noise is still an issue (its always an issue - 95% of all problem solving in audio is simply moving the problem somewhere else) but the importance of the soundcard is reduced.

Nice explanation, and by lowering bass signals when cutting the groove the groove width is decreased and so playing times increased as a bonus. Some speciality drum and bass LPs can still have as little as 16 minutes per side  and the 1955 "Blue Moods" Miles Davis LP was allegedly deliberately made short (under 27 minutes) to allow deeper bass.

admittedly nit picking: no cartridge can track full scale 20 kHz - maybe 5 kHz for exotic cartridges, typically 2-3 kHz "power bandwidth" is the max at which ordinary records were cut

that cuts 12 dB or more off the practical max dyanamic range required for "digital RIAA", to < 30 dB - as long as you don't demand "accurate" capture of mistracking

I still think digitizing a phono preamp's already EQ'd output is likely better, "safer"

Quote:

Originally Posted by Pitou 

Here is a test result from RMAA. I have no reference source, so I used the loop trick (connecting the line out to the line in of my sound card)

http://www.filefactory.com/file/r0usr0wosnh/n/_MME_Wave_mapper_24-bit_96_kHz_sav

What do you think?

It may be OK if the input is at a sufficient level and does not need further processing in software (i.e. it is already preamplified and RIAA corrected). But even a cheap sound card is a significant improvement over the A/D performance of typical onboard audio (which is still better than vinyl or cassettes nevertheless).

Quote:

Originally Posted by Pitou 

Is there a way to attach files here?

You can attach images, but not other types of files.