Click to go back to main page

pro audio f.a.q (frequently asked questions list)

Can digital audio reproduce sound with the same transparency, depth and quality as analog recordings?

A good 24-bit 96khz ADAC records audio with far more transparency, depth, and faithfulness to the original signal than is possible with any analog technology. Analog recordings often get described as having more 'depth and 3d stereo field' - these are actually properties inherent to the distortions produced by analog mediums (consoles, tape, preamps and electrical effects units) which add that 'extra something' to the clean signal (that 'extra something' can be distortion, harmonic distortion, eq colouration, frequency roll-off, crosstalk and a variety of other subtle colourations). We have gotten to the point with digital emulation where all of these effects - and more - can be reproduced in the digital domain, when desired - without some of the less pleasing distortions that analog technology introduces (hiss, unpleasant distortions). In short - we can get analog 'sound' when we want it - and not when we don't. Purists will say that the plugins can never sound 100% the same as analog componentry. And they're right, 90% of the time. But the gap is closing, and sometimes, those digital components sound better than their analog counterparts...

What is a console? Do you need one in a music studio?

Consoles, also known as 'desks' or 'mixing desks' are large pieces of equipment (typically analog but sometimes digital, or a mix of the two) which are sometimes used in studios to mix, monitor and record audio - analog consoles impart a sound on the recording process which is difficult to emulate, often much-desired for products like rock/pop recordings, and subtle mastering. They are the things you'll see in shots of studios with a whole bunch of volume sliders up and down them (which gives the impression of lots of things happening which aren't - since most of those sliders aren't used most of the time - but hey - it sure looks impressive, right?).

Up until recently, software developers were not able to exactly emulate the subtle analog effects (see above section on analog recordings) which consoles impart on the mixing process. Only in the past three years (prior 2010) have these systems been successfully emulated, with products such as Steve Slate Virtual Console and the AlexB Nebula consoles, giving (virtually) the same results as properly-tuned consoles worth upwards of $45000 USD, to about 90% accuracy. They also allow for the easy trialing of different consoles for any given recording, allowing much greater flexibility and quality of sound. Soul Studios has chosen to go down this route, rather than (a) shell out for a $45k console or (b) settle for a less than top-of-the-line console, and (c) being forced to use one console 'sound', all the time.

Incidentally the console sound is not suitable for all recordings. It introduces quite a large amount of upper frequency and extreme low frequency roll-off, harmonic distortion and crosstalk in signals, which is not suitable for - for example - orchestral recordings, and many other genres. Where it works best is rock and pop, ie. contemporary setups, where it gives an easier separation between the elements of the recording without sounding hollow, flat and anaemic, but it works less well in my experience for some electronic genres - where it can give an 'old' and somewhat muted sound.

For an understanding of how little recordings change between a virtual console and a real one, take a listen to the following examples:
The recordings were performed by steve slate himself - one is a recording processed through a real console, while the other one is the alpha version of the virtual console (needless to say it's improved since alpha stages). Can you pick which is which? email me your result and I'll tell you whether you're correct. For reference, here is the recording without any console processing.

Do all recording studios use Pro Tools?

According to online polls many studios are using other host software nowadays - including but not limited to, Logic Audio, Cubase, Reaper, Tracktion, Samplitude/Sequoia, and many others. There is still a sizable portion of studios using Pro Tools as well - Soul Studios is able to export projects to Pro Tools, Logic Audio, Cubase, and many of the others using AAtranslator, and uses Tracktion for most work. All of the main competitive products have equivalent or better processing than Pro Tools, however the major differences are in workflow. Personally I prefer the workflow of Tracktion: No big deal.

Is outboard hardware better than digital plugins?

Outboard equipment *can* be better than digital plugins - sometimes - it can also not be, depending on the quality of the hardware and the quality of the plugin in question. Personally I prefer the workflow and ease-of-use of good quality plugins, and find the sound quality to be comparable. Many people prefer hardware, but opinions are often divided as to whether the differences in perception between hardware and software are a result of personal preference, a product of different workflow, or objective sound quality differences. I would, largely, venture the former two rather than the latter one, in the present age at least.

Are U87/U47's, Shure SM7b's, and the other older mics the best microphones?

Despite the fact that many older mics sound very, very good (and will also cost you an arm and a leg), some newer mics are as good or better than their old counterparts. For example, a Rode NT1 (a beginner-level mic created in Australia) custom-modified by Michael Joly comes out better in recordings (as demonstrated via blind listening tests on various forums) than a U87. A newer mic, in the right hands, can sound better than an old one. Don't let names or their legacy fool you - listen to results, instead..

Soul Studios uses custom-modified Rode NTK's (similar sound to U87's, with a slight upper sheen), modified and stock Octava MC012's (general-purpose small-diaphragm condensor mics), Shure sm58's (the industry "standard") plus a variety of others. Finally I'd like to say the best mic is the one that sounds best in the particular scenario you're using it in - some are more general-purpose than others, but there are no 'definitive' mic's per se.

Does a control room needs to be completely separate from a recording room?

There are several good reasons to separate: isolating the control room environment for the recording engineer to hear the individual elements while recording, the reduction of incidental noise from the control room being picked up on the recording, and various other factors.

But there are also reasons NOT to separate: the physical separation between the performers and recording agents, more difficult or abstracted communication between those two, the inability of the engineer to hear exactly what the band is hearing. There are some prominent producers such as Brian Eno who prefer a single-room environment for artist communication and guidance reasons.

At Soul Studios I've opted for the best of both worlds - there are two recording rooms, one isolated, one not. The control room is separated, but not blocked from the first recording room- since my computers are custom-designed and produce less noise than a fly this allows for a clear separation of band space and control space, without the need to create artificial barriers between the recorder and performers and without any incidental control room noise. Then there is a secondary smaller sound-proofed recording room, which is completely separate and isolated from the control room. This is useful for when other incidental noises (for example, band-mates) become problematic or a performer feels 'exposed' in the larger, non-separated room.

Is mastering better done on software or hardware?

Mastering is firstly about ears, and secondarily about gear (the same logic actually applies to mixing as well, but whatever). If a plugin (or suite of plugins) is truly up to the task, then it's merely a question of whether the person in charge is capable of hearing the minute differences that can be, or need to be made, adjusting them in the most appropriate fashion using the most appropriate technique.

There are hardware tools and there are software tools, but personally I believe the software tools have far outstripped the hardware tools in the past five years (prior 2010), in terms of both quality and capabilities.

I've personally been using Izotope Ozone for about 9 years - mastering is a steep learning curve, and a subtle art, but it is definitely doable with a software-only approach. Recently I've started using Izotope Alloy in tandem with Ozone, for stem mastering, as well the digital console stuff mentioned earlier. Having compared my in-house masters to masters sent overseas to standalone 'mastering houses', I believe I'm equipped. Have a gander at the samples, make up your own mind.

Do you need X/Y brand convertors/DAC/ADACs to do a proper recording?

There are many good convertors in the world, and yes the X/Y/whatever's may be good, but they're not the only tool at disposal. I personally use a Black Lion Audio Modded 896HD with their custom-built MK2 wordclock, which is similar to (same internal convertor) but better than an Apogee Rosetta by virtue of the somewhat more exquisite wordclock. I am happy with this. At the end of the day, small tweaks like better convertors etc are only of any use if the engineer's listening and mixing/mastering skills are up to task. The best DAC in the world won't save a bad mix, but a good mix will be a good mix on anything (it just will sound better recorded through a good ADAC).

Why record at 96khz when humans can only hear up to around 20khz anyway?

The main reason to record at 96khz is not so you can capture higher 'ultrasonic' frequencies (though some would argue this is the case) but because the filter in the ADAC (the part which removes the frequencies above nyquist) does not have to be as steep when recording at 96khz (at 44khz it requires a steep rolloff between 20k-22k to adequately filter the signal) - this improves the quality and precision of higher frequencies in the audible range. Daniel Weiss's (an imminent audio hardware developer) interview has more details on the subject for those interested.

The other good reason to work at 96khz is that most properly-designed plugins work better at upwards of 60khz, as the error part of the signal processing can be pushed into the inaudible regions (upwards of 22k) rather than interfering with the audible signal.

And lastly - it just sounds better. (a bit of bias)

Why record at 24bit when CDs/DVDs/MP3s are 16-bit?

Because the increased amplitude resolution decreases aliasing (a form of digital degradation) of the signal when recording, and also because 24-bit audio has far greater headroom than 16-bit audio - this means we can record in a signal at a very low volume and not lose precision. With 16-bit recording you have to keep the signal loud or you lose amplitude precision, and you risk distorting the signal by doing so.

In fact, audio programs and plugins work at either 32-bit or 64-bit internally for these same reasons. The end result is that when you mix and master at higher amplitude resolutions, finally dithering down to 16-bit for the final product, you get a recording with more depth and spark than if you had recorded and processed at 16-bit.

Is it better to work at twice the product samplerate ie. for video record at 96khz, for CD/MP3 record at 88khz?

This comes from a misunderstanding of sample theory. The conversion process from 88khz to 44khz is much more complex and precise than simply (number of samples / 2). A larger number of samples simply means a more computationally-expensive conversion process, and a more accurate result (try and think in terms of the quality of resampling a 300dpi scan of an image down to 75dpi, as opposed to resampling a 120dpi scan of an image to 75dpi). Ergo - more samples in source = more accurate representation of original signal in destination (if the wordclock and AD convertors are up to the task). As stated above, unless you're trying to capture ultrasonic frequencies, any frequency rate upwards of 60khz is sufficient for digital audio recording, however I prefer 96khz to 88khz as it sounds better in the mix - this may be because more plugins have specific processing paths for 96khz than 88khz.

Is digital audio 'discrete' ie. does it break down an analog signal into a series of digital 'steps'?

While the resultant information is indeed a series of discrete 'stepped' samples, PCM audio is designed in such a way that - with some minor inconsequential (and largely inaudible) changes - the original continuous analog signal is able to be reproduced (minus the frequencies above nyquist) from these stepped samples, once run back through an adequate DAC. For more information Dave Lavry's PDF on the subject goes into more detail. In basic terms, the stepped signal which PCM is, in fact represents a continuous signal (though not flawlessly - no recording medium can claim to do that).

If you disagree with the above points and want to discuss feel free to get in touch!