increasing gain: hardware or software?

[hranis]

Hi,
I've been a happy Audacity user for almost a year now. I'm using it to digitise my LP collection. I have a PC with a Lynx L22 soundcard, which is not supported by Audacity (of course). The issue is that input signal from my preamp is fairly low (I'm using an MC cartridge and the phono stage has limited gain), so I've been using the 'amplify' effect to get the signal to the -3dB level that I can archive / play / burn / etc. Some records are so 'silent', that they need an additional gain of 17-18 dB.
My questions are:
1. Is this procedure giving me the best sound quality? (Admittedly, I'm very happy with it, but wonder if it can be made better).
2. Would I be better off providing gain upstream of the computer, on the input signal? This is probably a costly option, but if it means better sound quality, I will consider it.
3. Is there a difference between the amplify and normalise effects? Am I using the best digital option in this case?

Any comments and/or suggestions are very welcome.
Cheers,
DH

[waxcylinder]

3. Is there a difference between the amplify and normalise effects?

A spate of postings on the forum recently indicate that yes there is a difference between normalize and amplify - and this was unexpected behaviour that surprised us.

Normalize - appears to work independently on each stereo track - bringing each track in turn up to aspecified level. Note that this could potentially change the stereo image

Amplify - appear to apply the same specified amount of amplification to both stereo channels.

Note too that Normalize can be used to remove any DC offset - but this can be done with zero normalization, so you can remove DC and then use Amplify it that's what you choose.

I'll leave the amplification questions to the experts......

WC

[steve]

I've been using the 'amplify' effect to get the signal to the -3dB level that I can archive / play / burn / etc. Some records are so 'silent', that they need an additional gain of 17-18 dB.

So your signal level is around -20dB.
I would say that this is a little below the optimum, but not much. If you are happy with the sound quality I would not worry. I frequently record at a peak level of around -12dB and there is no discernible difference between recording at that level and pushing it up close to the max - it just means that I rarely overcook the recording into clipping.

Is it possible that your pre-amp was designed for a moving magnet cartridge? That would explain the low signal level. Is your pre-amp switchable between MM and MC (I guess you've already looked for that option).

Adding an extra hardware gain stage will introduce additional noise and distortion, so there will probably be no overall improvement. Changing your pre-amp for a good quality MC pre-amp (with more gain) may give better performance, but it will also depend on the quality of your other equipment as to whether it is worth the upgrade.

[hranis]

Thanks for the comments so far.
It is interesting to know the difference between amplify and normalise - I am very happy that I've been using the amplify option, since normalisation tends to ruin the intended balance of the recording. As I said earlier, I am very happy with the final sound quality, which apart from volume does not have any audible difference as compared with the original. But then, I am recording in hi rez (24/96) and playing the tracks it back straight from the hard drive.

With respect to the second comment, my pre amp has enough gain, even though the MC cartridge has a failry low ouput (0.4mV). What I believe is happening, is that the sound card has a small 'step down' device at its front end, reducing the signal strength by about 10dB - I am led to believe that this is their way of preventing clipping. If I could somehow bypass this section, I think that all would be OK.

The main reason I started this thread, however, is to hear opinions from similar users on the question of hardware vs software amplification, and the sonic benefits (or otherwise) of each method. I am particularly interested not what happens at the top end of the sound signal, but at the bottom - how is noise treated and which method would give a 'cleaner' sound, without applying any effects, of course.

Any thoughts?

[kozikowski]

<<<Any thoughts?>>>

Stop using a PC? Nobody is going to throw awards at PC sound card makers for their high quality and superior performance. Something about nobody willing to paying more than $12 for a sound card.

On the other hand, even though Macs will not handle microphone level, their analog line-level input is excellent. That, a good turntable, preamp, and Audacity and you pretty much got it licked.

Koz

[waxcylinder]

Stop using a PC? Nobody is going to throw awards at PC sound card makers for their high quality and superior performance.
Koz

Or buy an external soundcard - mine works a treat on both my PC's desktop and laptop.

WC

[waxcylinder]

The main reason I started this thread, however, is to hear opinions from similar users on the question of hardware vs software amplification, and the sonic benefits (or otherwise) of each method.

Ok I know I said I'd leave it to the experts .... but heres my two penn'orth anyway.

When you do the amplification in software, as well as amplifying the required signal you will also be amplifying the noise floor of the deck/amp you are using - not really a problem if you have a very low noise floor on your kit. Listening tests pre and post amplification will be your guide here. If your pre-amp will not deliver a high enough signal level for you then this is the only route without buying additional hardware.

In hardware you are dependent on the qulailty of the pre-amplification circuitry and the noise floor that you get with the kit.

I actually use a combination of the two: my ART preamp has a gain control and a useful LED indicator to show when I am getting into hardware clipping. I tend to set this to its max level prior to recording (note this can vary with each LP and needs to be reset carefully for each album) - the signal level delivered to Audacity I then control using the gain control on my Edirol soundcard. This gives a good strong signal for recording and normally rquires no further amplification - though sometimes after clickremoval I do use a little boost.

When I record FM off-air I tend to set the Edirol with plenty of headroom - as many of these shows are recorded when I am not there. I mormally lift these a little with Audacity's software amplification.

Both produce excellent results.

But as always, your true guide is your ears (with good loudspeakers or good headphones). It's why recording studios have big and very detailed loudspeakers.

WC

[steve]

Stop using a PC? Nobody is going to throw awards at PC sound card makers for their high quality and superior performance. Something about nobody willing to paying more than $12 for a sound card.

PC's are capable of extremely high quality sound and can out perform a standard Mac audio interface. However, not without using a high quality sound card (either internal or external).

The performance of "on board" sound cards (built into the mother board) is usually pretty poor on PC's and particularly so on PC laptops. In the intense competition for high spec/low price (not a competition that Macs take part in), the included audio device is one place that manufacturers cut corners. Adding a sound card that is designed for music will push up the sound quality to a very high level.

Internal PCI sound cards for full size computers start at around £50 (GBP) and can give you true (hardware) 24/96 performance. External USB audio devices are suitable for full size computers or laptops and start at around £25. PCMCIA Audio Interfaces for laptops start at around £125.

There are a couple of disadvantages of using external sound cards. An internal sound card will usually be on and available as soon as you switch on your computer, whereas USB audio devices need to be recognised by the operating system after boot up. Internal sound cards are usually configurable so that the output can be routed directly to the input (for recording sounds that are playing on your computer) but is not usually possible on USB devices without using a loop-back connector to physically connect the output to the input. With USB audio devices it is often not possible to monitor the input and the output simultaneously - this is not a problem if you are recording from a turntable or a mixing desk, but may be a problem if you are recording directly from a microphone and monitoring through headphones. USB audio devices will typically have higher latency than internal devices, but this only really becomes an issue with software studios when virtual instruments are being used, and not for basic recording. Many USB sound cards will provide zero latency monitoring of the input through a headphone socket on the device.

Firewire devices are popular for multi-channel recording and can offer high performance, low latency, multiple input and output channels, and high audio quality. In effect they offer the best of both worlds (internal and external sound cards) but generally at a higher price. This is the most common solution for professional recording studios.

External devices (USB and Firewire) will often be compatible with both Mac and PC, though you need to check the specification before buying.

For PC's, my preference is for PCI sound cards as they provide high quality, ease of use, convenience, low latency and, versatility at affordable prices.

A Mac is very likely to give better sound quality than a normal consumer PC, but upgrading the sound card on a PC will bring the audio performance up at a fraction of the price of a Mac.

Note that price alone is not the best guide for choosing a sound card. Many sound cards are optimised for games or video rather than sound recording. Common brand names for music sound cards include; M-Audio, EMU, RME, MOTU, Apogee, Swissonic, Lynx, Sonic Core, ESI, Tascam, Alesis, Steinberg, Presonus, Digidesign, Mackie, Edirol, Behringer, Line 6, Echo....
More expensive sound cards will often include special features for studio use, such as a highly accurate clock, sync to external device, hardware DSP, hardware mixing, ultra low latency.... but such features are not generally required for basic recording set ups.

[MIgartua]

Has anyone tried recording using the Mackie Onyx 1200F external sound card?