Page 2 of 2

PostPosted: Mon Sep 05, 2005 4:34 pm
by citguy
Although "headphone out" is usually in parallel with "line out" it many times will have different impedence and consequently different sonic characteristics from tone controls. My old Dynaco has a choice of "line outs", one that is controllable by tone controls, volume etc and one that is not. Also the manual explains some controls such as "loundness" will have less effect on the 'headphone out' than the other outputs. The use of headphones can be more efficient with error correction on WC, so it is important to feed the soundcard with a 'line out' that is not disabled when headphones are plugged in. I have not considered the fact that some users may be using an output from their sound card to monitor the input signal. I monitor from the control preamp.


PostPosted: Mon Sep 05, 2005 5:26 pm
by Glenn
Hi dead :wink:;
First of all, welcome to the fold. You wrote:
As indicated in earlier posts the best solution is to use a lower level analogue signal, which avoids the digital clipping. You may be lucky enough to have an adjustable line out on your pre-amp or amplifier. Unfortunately this isn't always the case. If you are electronically minded I guess you could knock up a potential divider circuit to do the job.

Alternatively there are some phono pre-amps on the market which do include the necessary output level control. These can be expensive. I'm taking the opportunity to recommend one relatively cheap pre-amp that does the job for me. (Another approach might be to buy a deck with an in-built phono pre-amp. These can be quite cheap. Bush make one for about £50, but I guess you get what you pay for.)

In the instances where one has no means of controlling the volume level on their stereo equipment (i.e. monitor-out), an alternative would be to use a portable stereo with line-in and headphone-out as an intermediary volume control. These can be had quite cheaply, but will likely reduce the fidelity somewhat. Some people connect their record player directly to the pc and use software to apply the riaa equalization, but this approach is prone to suffer line noises because of the extreme signal boost which must be applied to normalize the recording.
As you say, I agree that the best approach is to use a phono preamp with a volume control at the output. As mine has a continuously-variable attenuator, I set the line-in record level to maximum and adjust the attenuator to the highest level possible without clipping. This approach will also yield good results where the only available output is through a receiver's headphone jack.

Analog signal clipping

PostPosted: Mon Sep 05, 2005 7:26 pm
by deadpoirotsketch
Sir, I am glad your Audigy doesn't display this problem. (BTW which version of the Audigy?).

The audiophile 2496 does display the digital clipping problem - here is a reference to a thread at hydrogenaudio forums if you are interested. As the 2496 card is very popular with people transfering vinyl to CDR it is worth mentioning here. ... =12349&hl=

I believe it is quite possible that the variance in design of different makes of soundcard might include variance in the ability to handle different levels of "line level" input. Knowledge of such might be useful in informing any soundcard purchasing decisions - however I for one never seem to find out about drawbacks with soundcards - such as the internal resmpling issues in Audigy cards - until I have bought one on the strength of wonderful reviews. Yes, I have an Audigy sitting alongside my Audiophile 2496 and I personally prefer the results from the 2496, but the only thing that matters in the end is what each of us individually hears and finds good.

If anyone is experiencing digital clipping with their particular soundcard then I hope my earler post helped - that was my only intention and I didn't make the claim that ALL soundcards would exhibit digital clipping, or that it couldn't be rectified in software.

Having said that, I find it hard to visualise how software could control the level of a line input to avoid digital clipping *before* it has been through an ADC and entered the digital domain. I am not an electronics guru and I would welcome someone of greater knowledge explaining this in a way that won't make my head explode or simply saying you move the slider in the on-screen mixer!

I don't want to take part in a my soundcard is better than your soundcard argument, but I would like to understand better why some soundcards may be more vulnerable to digital clipping than others.

PostPosted: Tue Sep 06, 2005 4:10 pm
by Glenn
I'm not a soundcard expert but I believe it is possible that some cards could have a digital potentiometer on the analogue side of the ad converter that could be controlled with a software slider. This would be a convenience feature for the consumer as these digpots are known to have fidelity issues, and is possibly why the 2496 doesn't have them. I'm just guessing really. Of coarse, clipping is only a problem if it occurs. My software recorder doesn't well handle the quantization errors that result with clipping, resulting in high frequency polarity reversals that sound like loud pops. I use a peak limiter featured in the recorder that eliminates this issue. It's action doesn't prevent clipping in the da converter, but corrects the problem before it is written to wav. In most instances, it's effect is inaudible as I set the record level so that the clipped transients are few and far between, most often occurring during loud clicks and pops in the vinyl. The rest is pure, sonic joy.

PostPosted: Wed Sep 07, 2005 9:48 pm
by deadpoirotsketch
Hi Glenn,

I have to admit I don't record into Wave Corrector, rather I use quite a good sound editor - but I haven't a clue whether it or wave corrector contain any automatic software mechanism to "soften" clipping or other artifacts should they occur. I think we would probably agreee it is best to avoid the need to correct the waveform being recorded in the first place if we can. (Smile).

I think that clipping arises if the maximum voltage difference of the waveform peaks exceeds the fixed voltage difference at the card input. So in my admittedly naive way I guess the audiophile 2496 may have a smaller pd on the analogue input line than the audigy ( hope I'm using the right terminology). I really don't know if it is possible for that analogue input potential difference to be controlled digitally - i.e in anaology to widen or narrow the stable door. If any of my conjecture is correct it seems more straightforward to design the card with a large pd in the first place. As the audiophile 2496 is sold on the basis of digitising your record collection it does seem like a horrible fault to have. Seems to me that the maximum acceptable phono cartridge output should be highlighted if the manufacturer is not going to alter the circuit.

Luckily I haven't spoilt any recordings because of the control in my pre-amp. Much more luck than genius I am very ready to admit.

I wonder why record deck manufacturers don't go the whole way and build a preamp and good ADC convertor into a good quality deck and give us outputs that can go to a digital amp or to the SPDIF on a soundcard? Or even just sell a pre-amp and ADC in a box especially for digitisation of vinyl and shellac?

You may be right about a "digital potentiometer". Apparently Magix Audio Cleaning Lab v10 can be used without a pre-amp. You just plug your cartridge straight into the soundcard. I don't have any personal experience of how good this is. I guess the necessary RIAA EQ correction and gain are all applied in software - whether this happens before the waveform is digitised or afterward I don't know. If it works well then maybe we could ask Derek to copy the idea! (Grin).

PostPosted: Wed Sep 07, 2005 10:30 pm
by adaywayne
I think that clipping arises if the maximum voltage difference of the waveform peaks exceeds the fixed voltage difference at the card input. So in my admittedly naive way I guess the audiophile 2496 may have a smaller pd on the analogue input line than the audigy ( hope I'm using the right terminology). I really don't know if it is possible for that analogue input potential difference to be controlled digitally - i.e in anaology to widen or narrow the stable door. If any of my conject
Clipping will occur if the input signal (analog) to the soundcard exceeds 0.775 volts (+ or -) which, in consumer electronics convention, is equal to 0dB. However. Some the A/D converter of some soundcards, will give significant distortion below this level, which is why a max input of -5 to -3dB is always what I aim for. If I am wrong about this, I'm sure someone will correct me.

After saving the recording to your hard-drive in wave format, it is always a good idea to scan through it with Wave Corrector or other software to visually check for any clipping before burning to disc.

Fader circuit

PostPosted: Wed Sep 21, 2005 11:02 pm
by deadpoirotsketch
The Audiophile 2496 has an input impedance of 10kilohm (according to the manual). If anyone wants to construct a pot-in-a-box to reduce the analogue line-in signal to this sound card then you can use the following:

10 kilohm, linear taper, dual-gang, 4 watt potentiometer.

Typical UK supplier Maplin at £1.99

By the time you've bought a metal box, dual core screened cable, phono plugs and whatever line socket you need to connect to your pre-amp or amplifier line-out cable, it will cost approx £8-£10. For the sake of completeness I added an earth wire and tag to attach to my PC chassis from the metal box in case of hum, and to try not to add significantly to the noise floor - but this may be overkill.

There is an interesting related article including wiring details here:
"Nine ways to adjust signal level"