news category created 4 July 2014 written by Dan Cox
Defining the principle outcomes for a high definition audio format:
I’ve been running a HD store for over a year now. It was born out of my approach to mastering, where I mainly focused in the analogue domain and a client need to deliver albums in a HD format thus being able to sell alongside their other delivery via Vinyl / CD / Lossy formats but with a distinct difference.
I’ve always thought the use of excessive digital limiting as one of the largest issues affecting mastering practise over the last 15 years. There’s now almost a generic use of digital limiting for the sake of perceived loudness in a majority of commercial delivery. It’s inherently expected by the consumer and client but one aspect to this is why not! If they want it, feel they need it, why shouldn’t they? Who’s to say what’s correct, or change habits of a generation? What we can do as audio professionals and music enthusiast is offer an effective but desirable alternative.
This is the key aspect to a HD delivery format; we need to make a clearly defined and consumer comprehendible product in the same way as Vinyl / CD / lossy has captured the public’s attention, irrelevant of the right and wrongs of the audio compromise.
We all know 24 bit sounds better than 16 and higher samplerate contain more detail but this is only one aspect of the compromise in current lossless audio delivery at 16bit 44.1khz. Apart from the bit depth reduction / dither / samplerate conversion, which we all know compromises the audio in the rate reduction during the mastering process, there is also the aspect of limiting for the sake of perceived loudness.
Mastering mainly in the analogue domain as part of the transfer path and from gain structure point of view, I’m not concerned with reducing the peak level unless there’s an aspect of the mix that needs corrective processing (i.e. it’s not been effectively controlled at the mix stage). The focus is always on the RMS with the aim of achieve excellent tonal balance and dynamic control, thus an effective master that will translate on all playback systems.
At no point in this am I concerned with reducing peaks for perceived loudness. The peak to RMS ratio generally finds a natural level depending on the genre and inherent dynamic of the music. Captured back to digital this ‘Mastered sound’ is, in my opinion, the best the music ever sounds. I should also say I may apply some form of digital dynamic manipulation after conversion at this point if required, but we now have a master with effective Dynamic Range. Obviously we have to obverse the peak level on conversion back to achieve the best overall level during conversion.
In general because I’m affecting tone and dynamic in the analogue domain I’d capture back at 24bit 88 to 192khz dependant on source. I nearly always double the source rate to retain more of the analogue transfer path capture. To note, there’s no point recording or delivering HD at 32bit when we’re all using 24bit for AD DA. We just need to make sure gain structure is correct. Thus, we now have a High Definition Full Dynamic Range Master, which. I’ve been referring to as FDRHD, hence the digital download store name http://www.fdrhd.com/ . Effectively, this delivery is how many engineers used to work (and probably still do) before the days of digital, purely in the analogue domain where we were never concerned with peak reduction.
I’d take this FDRHD and digitally limit to perceived commercial level, compensate for tonal shift and rate reduction to achieve a ‘CD Master’. See how this translates to AC3 and compensate again for this aspect of the project’s parts production. (Delivering for ‘Mastered for itunes’ but as pointed out by a previous post this is all really about apple building up a HD catalogue and not improving audio delivery!) All of these points are compromising the original ‘best sound’ achieved. In this, I’m not saying I don’t work with the limiter in-line on the end of the transfer path to achieve the ‘best sound’, I do for the CD output, but primarily the initial focus is always on achieving a fully dynamic / solid sound. The rest is always a compromise, though not massive, there is a distinct difference.
Overall from this what I find very interesting is there’s more of a change to the positive in sound from the lack of limiting for loudness than the HD rate reductions. (Basically we retain transients, the stuff mix engineers spend ages putting in!) Both aspects combined; there’s a clear difference a consumer can hear on a quality system. The sound many people are used to from the days of vinyl before the invention of digital limiting but without the physical restrictions of vinyl. (I supply whenever possible an FDRHD version to the cutting engineer as it always translates better than the limited version.)
So; I think there’s more than one factor in defining a new HD audio format.
There’s the audio aspects:
1. The file format shouldn’t matter as long as it’s lossless, whether it’s FLAC, WAV, AIF, DSD, ALAC or whatever the future holds. It just needs to be at the highest rates possible the project and systems used can practically allow.
2. The audio resolution should be as high as possible taking in to account the restriction of the source and mastering environment but an assumption would be a minimum bit depth of 24 for PCM formats.
3. The audio shouldn’t have any limiting for perceived loudness. As mastering engineers, we all know what that means. (Use as many limiters at you like but only for the sake of making it sound better! Not just louder!)
4. It should be open to everyone to make, restrictions only lead to compromise.
And the commercial aspect:
1. Having a distinctly different format helps a label / artist target a different audiophile market place. In the same way they’d make a vinyl alongside CD and lossy download. It doesn’t affect the other markets – just enhances the product profile overall.
2. It’s needs to be clearly a different product to make it easy for the consumers understand.
3. It’s needs to be an open format, anyone can sell or utilise if it fits the simple audio requirements.
Other general aspects I think are useful in helping to deliver a new product:
1. Album download only, no single track downloads, which gives value to the concept of an album in the same way as vinyl always has.
2. Full credit listing for all involved in the creation of – something the MPG has been focused on for a while. (Much can be achieved with metadata in today’s files formats.)
3. All the above defines it as a high value product, hence commands a high price and thus delivering much need revenue / value back in to what many would see and feel as our currently devalued industry.
To sum up, I’ve been actively engaging in this for some time; you can read more and listen at FDRHD.com, there’s some free downloads available. Please don’t be judging the workings of the site or audio as such. It’s the concept of defining a format I would like to promote an open discussion upon. Your thoughts and professional feedback would be very welcome. I’ve no agenda with this project other than to help people hear better audio.
JP Braddock – Mastering Engineer – Formationaudio
PS – feel free to contact me directly to discuss.