Mastering Without Clipping Converters

[Paul Frindle]

Originally Posted by MattGray

Yeah understood, the point of mentioning the kick & snare scenario was only to say that limiting or clipping program material is a totally different kettle of fish to processing sine waves. It's probably why there is so many different options available. Because there is far more variables to consider with every song/mix & you're right in saying that there is no 'single' solution or approach to handle every situation effectively. What sounds good or works for fast transients & drums won't work on sustained peaks or certain instruments as I alluded to in my initial post.

Generally speaking you are right. If the clipping and limiting is of a conventional type the appropriateness of either/both will depend on the dynamic content of the music and in particular where in the freq spectrum the peaks you are trying to lose actually are - and how long they last.

Also it will depend on whether the sound of the clipping can be a believable part of the instrument itself. For instance clipping percussion has long been used to create hardness on otherwise over-soft sounding snares and kick drums - it can be a life saver for the mix engineer faced with badly recorded material, or where production staff want to change emphasis ofter the event..

But given that all sounds can be represented as the sum of sines over time, magnitude and phase, it is actually possible to create a distorting process that will act favourably with transients as well as continuous sounds.

If I get time I'll post another file with a snare or some such.

I'm not sure if I'm hearing the 'softness' as harmonic distortion & whenever I've tried to use DSM it's always been with the limiter out & with a longer attack time at least 30-50ms with a release suitable for the frequency range I'm trying to control. I do find that a release time of 200ms can be a bit restrictive & often too long when dealing with peaks, so maybe that may also contribute in small part too the 'smoothing' effect I'm hearing - not sure. The thing is that it's all so interactive even the slightest changes can change things sonically for the worse if you're not careful. Even when you've got it working to fix something on a mix it still comes at a slight sonic penalty to my ears but to be fair I've never been a fan of multiband or most linear phase processing for similar reasons as well. However, occasionally there is exceptions to this rule.

I think there are a couple of things here. Firstly because the DSM has so many bands in comparison to conventional multi-bands it can act on tight individual bands on the MF and HF region (where many transients lie) - whilst leaving the bulk believably intact. This, plus the fact that it is working with the underlying spectrum capture of the music, means it's quite easy to be lulled into initially accepting the 'average' sound (because it's so believable) to a degree where you can start ignoring the transients. Then later you can listen to stuff and have a double take about what you missed before? It's happened to me a few times - for some purposes the overall compression capabilites can almost be 'too good' - and the limiter section is needed to dirty it up - if you know what I mean.

In many ways you have to think about the dynamic freq content rather than a conventional notion of limiting. It's a very sharp tool that can cause dramatic changes.

Secondly I completely agree about the linear phase stuff. It's unnatural in the real world, spreads stuff in the time domain and softens things by requiring that the impulse response must start before the transients that caused it arrive. However beware of dogma here - the highly acclaimed Weiss unit you like uses linear phase filtering!!

Well yeah the Weiss hardware although being digital does tend to hold it's 'worth' a lot more than any plug-in. Whether that's purely because it's hardware or because it just works better than any plugs I've heard I can't say. I guess one of the advantages is that it has a significant amount of dedicated DSP so it's less likely to be sonically 'impaired' at the sake of resources which I'm sure a lot of plug-in designers have to be aware of & struggle with to keep that balance between sonic quality & CPU resources. Whatever the reason, I've yet to hear a plug-in that compares 'sonically' to the Weiss DS-1 Mk3 in terms of transparency even though the original design is years old. Daniel Weiss obviously knows what he's doing, Alexey Lukin is another who does transparency in DSP extremely well.

The use of resources is an interesting and rather odd situation. In fact the average $500 machine from your local store has many times the processing power of a great many of the H/W processing boxes out there. There are even some expensive audio processing extention cards/boxes that won't even run a single DSM - when a Mac Pro laptop will run hundreds and still play and mix 24 tracks of audio in real time!! It will not be the dedicated processing that gives a sonic advantage.

However Daniel Weiss does indeed know what he's doing to a very high degree :-) This represents any advantage he has - not the box itself IMVHO :-)

One other aspect of a hardware box is that the manu is not slave to continual operating system updates/changes - which means that unlike plug-ins you are not forced to update it every year or so. This is a big overhead for plug-in manus - and it degrades the value of their products to an extent where some are having to charge for the periodical updates. This is highly unpopular amongst users - understandably..

The other advantage of a HW box is that as a manu you are very usefully disconnected from the vagaries of the multitude of host S/W and platform issues that plague plug-in manus with support issues and forced updates. This is a dramatic overhead to S/W only products.

But of course the single major advantage of plug-ins for recording and mixing (apart from radically reduced cost) is the integration with their work environment right within the W/S app itself. This is an extremely important creative and artistic advantage that H/W boxes miss out on. But as things are, despite the great advantage of S/W apps running on cheap H/W, making plug-ins is not a by and large a profitable business - not least of all because they are viewed as second rate by the users - and the whole world is obsessed with emulation of other people's stuff, from now and the distant past. It's almost as though - even in this world increasingly set free from physical constraints - if something hasn't ever existed in the form of a real 'box' it's automatically viewed as something of lesser 'value' - even though of course the music itself we are all trying to create has always been a non-physical thing?

Addtionally, for the mastering fraternity expense and exclusivity of equipment is often an important commercial factor in a world more obsessed with 'visible value' rather than the engineer's valuable experience and skill - which is what they ought to be paying for? I can understand mastering engineers being a bit reticent in admitting to using plug-ins in this current market..

Of course, I actually prefer original DSP rather than those that claim to model or emulate analog hardware. There is some indispensable digital tools that I use regularly in combination with some great analog hardware. I think both have their place in a mastering chain & have their own unique strengths & weaknesses. Together they complete the ideal workflow & sonic signature for me.

Absolutely - there is no single magic bullet app that can do all you would ever want (as yet). People must grab and make use of whatever they have available to do their art - without any slavery to dogma :-)

It reminds me of an argument I once had with someone complaining about the lack of technical 'accomplishment' and instrument hierarchy in popular music, basically saying that anything later than Beethoven could not be described as music at all! I replied that as far as I was concerned I'm happy for people to bang dustbin lids together - if it's artistic and makes me feel something I can identify with emotionally!! I personally don't think there's a valid place for snobbery anywhere in the artistic process.

But the main artistic aim of integrated digital processing is to get the creative process under your control, rather than tha phase of the moon - or how worn your tubes are. I want to release people to do their art and not be slaves to the unknown :-) Doing this is absolutely not to delete the past at all - it's taking the past with us into the future in a coherant progression.

Sorry for the long post..

[macc]

Thanks for your response to my question, Mr Frindle

[Odeon-Mastering]

Originally Posted by Paul Frindle

Sorry for the long post..

Many Thanks for the long posts.

[LD Productions]

alot of good info in this post! when speaking of snares or certain transient sounds in particular, it is interesting to me that some sounds maintain punch much better than others when clipped (or limited). my advice to mix engineers would be to clip your mix like crazy before sending out to an ME as a test to see if it will hold up. obviously dont send the ME that clipped mix, but rather use it as an audition. if you know the ME is gonna slam your mix, use that to your benefit. i usually take the crappiest ITB limiter i have and destroy it to get a -9 rms mix, and listen if my center channel goes haywire, or still sounds somewhat normal. off topic a little, but i get alot of projects to mix with drum samples from libraries (i will not name names, but im sure you can guess what libraries are popular these days), and the sampled snare sounds NEVER stand up to any limiting very well. even a poorly tracked snare with a 57 and terrible hit consistency seems to "hit" through much better with clipping or limiting, if saved in a good mix with the right eq and compression. sorry, WAY off topic...

[Table Of Tone]

Yeah, cheers Paul!

[MattGray]

Yes thanks Paul for your time, informative & educational posts.. It's always good to hear from a seasoned DSP developer such as yourself.