Skip to main content

I have been asked to master a few recordings for a friend, and some of the tracks have haze of digititus (for lack a better word) over them.

My question is: what are some ways you have used to combat this problem?

Here is what I tried in this particular solution:

I ran the problem tracks through my Crane Song hedd 192 and applied some the tape saturation effect. This is kind of like, covering the tune is warm fuzzy blanket. It only masks the problem, it doesn’t correct it.

I then applied some very gentle (1dB or less) EQ cuts. This seemed to reduce particularly nasty frequencies that where sticking out of the fuzzy blanket a bit too much.

I am really interesting in hearing your approaches to the problem

Thanks!

Comments

Massive Mastering Thu, 02/28/2008 - 16:13

That's the problem with digititis (tis, tus, however).... It's really something that has to be handled at the source and during tracking.

And if it means anything, the only time I see (hear) it a lot anymore is when people track too hot. Which is unfortunately common in the "rookie" ranks... But I digress.

Yeah, the HEDD's tape (and pentode) functions are nice for putting a little "cream" over the bitterness. Keeping with Crane Song, the Ki mode on the STC8 compressor is pretty great for that also. I'm not even above whacking a reel of 1/2" when some really nasty stuff comes in (as much as I'd rather not usually). The (Manley) Vari-Mu is also a pretty stellar performer for such things if you're careful...

It's definitely a "per-unit" process though -- Sometimes a whisker of EQ takes care of it right off the bat. Other times, tape emulation (or actual tape) helps in a more friendly manner. Other times, nothing helps.

Thomas W. Bethel Fri, 02/29/2008 - 05:23

I have to go with Massive Mastering's response.

Most cases of Digititus are caused by over recording.

One thing that I have found over the years is to use a declipper and or a declicker to get rid of SOME of the problems. You have to be very careful that you do not attack the attacks from the percussion but used judicially it can help GREATLY.

I use the built in declicker in WL and the declipper from the Sony Noise Reduction plug in package,

Sometimes putting a lot of analog type syrup over the Digititus just makes things worse.

cerberus Sat, 03/08/2008 - 12:40

hello, this is my first post here.

i do not think that digital clipping is the same as digititus.
imo, these are the correct definitions that are relevant:

digital clipping can only occur from overloading a fixed point
data bus. this would generally be from operator error; a
tracking/mixing/mastering engineer's mistake. it is a
"high level" effect, since it tends to occur at the
highest amplitude, or loudest parts of a signal
(or if embedded, were at one time).

but imo, digititus is a low level effect which can be generated in tools from internal
math equations that make low level errors. typical examples might include:
processes which produces aliasing; and processes which truncate or
round a signal to a lower bit depth rather than dithering it.

jeff dinces

Cucco Tue, 03/11/2008 - 10:51

Massive Mastering wrote: The (Manley) Vari-Mu is also a pretty stellar performer for such things if you're careful...

Very much agreed - but a little bit goes a VERY long way with the VariMu.

I would tend to agree with the assessments of digititis by Jeff and John. Digital clipping is a totally different beast and the best and (IMO) the only place to combat it is in the mix and tracking phase. Using EQ has always left me feeling dirty and naked when listening back to the track - specifically when using it to try to *warm* things up a bit (I hate to use this over-used term, but hey...)

Digititis comes from many places - poor microphone choice (over use of condensers - especially the lovely chinese ones that have a nice bump of 10dB or so at 5kHz), poor preamp choice especially when combined with the poor microphone choicie (imagine a song made of 20 tracks of Studio Projects C1s through a Mackie 1202...ouch - ear bleed time).

Another one comes from poor AD converters - although, this is becoming MUCH less of a problem nowadays as even modest converters in cheap interfaces are sounding pretty friggin nice nowadays.

Here's the biggest source of digititis IMO - overuse/abuse of EQ. We all know that EQ can introduce nasty phasing and timing issues. The place where this is most noticable and effective is at the top of the critical midrange - 2kHz-4kHz. When things start getting hashy up there, it really reaks havoc on our ears.

I recently did a mixing job for a guy in Indiana who recorded his stuff at a really nice studio with a really nice desk and everything. The stuff sounded great - but ONLY after I had him send me the un-effected tracks - that is to say only after the EQ was removed off of all (yes - every single one) of his 60+ tracks.

The recording engineer put EQ on EVERYTHING and while it made some individual tracks sound great, when they were all put together, it sounded just horrible. After all the EQs were removed, I think I used EQ on maybe 5 tracks (including 2-buses) in total along with some selective compression on other tracks (for example - a slightly fast compression with long sustain on the bass to remove the tubbiness - something that the recording engineer tried to do with a 50Hz cut of 3dB and a 2kHz boost of around 3dB) and the mix sounded great (of course, in my humble opinion.)

Anyway - just thought I'd weigh in...

Cheers-
J

cerberus Tue, 03/11/2008 - 11:28

Link555 wrote: Hey Cerberus, thats what I was wondering. I have noticed some ADC's sound much brighter in the 2-4k range. For exampe M-audio Delta 44 has a very aggressive midrange. Is this digititus?

imo: no evidence of digititus there: an analog
device or process may produce a similar effect.

here is a case where, imo: test tones/sweeps and a visual display could be helpful.
then one could try to measure the actual frequency response.
(maybe for converters this could get expensive, analysis in the
analog domain...)

i think your observation is so far: a psychoacoustic one. we can't be sure
whether the frequency response of the signal was altered yet.

e.g. it is a well known psychoacoustic phenomena that adding
hf noise to a signal (such as tape hiss) could fool us into
a perception that these frequencies are present in
the signal. or seem to bump the freq response
when in fact this noise is not correlated to
the signal at all. imo, that would not be
digititus; imo it would be: "noise".

jeff dinces

cerberus Tue, 03/11/2008 - 12:18

hi jeremy; i do not like equalizers. but it is a tradition in
mastering that eq is needed to make the riaa curve.

i've been experimenting with time domain effects, this is not an easy
way to master, and i cannot teach it. every job ends up with a
bespoke parallel chain that i feel is overly complex. such is
the cost of educating oneself while also earning a living. i
must become more efficient. the temptation to reach for
an eq... sometimes i feel i should just buy a sontec and
be done with it. one twist of a knob, and ... bingo.
(or so it is claimed).

jeff dinces

Cucco Tue, 03/11/2008 - 12:26

Oh, no, I completely agree that EQs are required and useful in mastering and tracking. It's the misuse or overuse that bothers me.

Also, when I try to compensate for a problem in the mix by using EQ, it has a tendancy to remain a problem or create new ones. (Of course, adding the traditional "air band" or making minor changes during mastering is not what I'm referring to. I'm talking about extreme cases - massive bloating at 80Hz-50Hz due to poor monitoring, extreme stridency between 2k and 7k, etc.)

I'm a big fan of certain EQs - I love the Crane Song Ibis (thinking about getting another) and the Weiss (can't afford one of those for a while - I love their dynamic LP EQ though!)

cerberus Tue, 03/11/2008 - 12:35

isn't a "dynamic" eq = a [digital] multi-band compressor?
if that is the case, i don't think we disagree too much.
john(massive) may not concur, however.

imo also: multiband is more or less the same as parallel sidechained
compression. i don't see why anyone should dislike it on principle.

jeff dinces

Cucco Tue, 03/11/2008 - 12:41

Yes.

And in fact, I've been finding myself using my UAD Multi-band compressor that way to a great degree of success.

I don't find it doing the traditional "compression" jobs.

My beef with MB Compression is a completely different beef in that most people really have no clue why or how to use it - just that they "MUST" to be a mastering engineer (of course referring to basement and garage engineers here...)

Where I've been having some success with it is in the higher frequencies where at too great of an amplitude, they can get strident but too low of an amplitude and they get lost. Of course, a VERY little bit goes a VERY long way.