Mastering - analogue vs digital and peaks MASTERING GUYS

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

tardishead

Well-known member
Joined
Aug 11, 2004
Messages
623
I am analogue crazy and have been running a vintage gear studio for more than 15 years. I have always left mastering up to the pros and just kept to tracking and mixing but recently I have been getting a lot of requests for mastering demos or lower budget releases. And they ask for this because of the vintage equipment that I own.
Some of the tunes I get asked to master are not recorded to tape and some are. The ones that are not recorded to tape always have a lot more HF peaks which even my fastest analogue compressors cannot deal with. And in fact because of that the peaks get bigger after compression or limiting because they are missed. So consequently this results in less dynamic range.
The ones from tape I never have this problem - they always benefit from medium to large amounts of compression or limiting and they smooth out beautiful.
I have been resorting to adding compression/limiting and then going to tape (AMPEX 350) which works well for some things but for others there is too much coloration and change of frequency response.
Is the only real way to deal with these peaks involved with direct to DAW recording to use PLUGIN digital compression/limiting?? I would rather take the analogue route but it seems my analogue outboard cannot cope.
My fastest limiter is 0.2ms attack which corresponds to a whole cycle of 5khz. Or 2 cycles of 10khz.

 
The big deal with mastering these days is to have a combination of digital and analogue processing. There are tons of turtorials or other info spread on THE web. Google is your best friend.
 
uh yeh thanks for that. Actually I am looking for a discussion on whether taming these kind of digital peaks can be achieved through analogue - or whether that is just the nature of DAW recordings that they contain these nasty peaks and they have to be dealt with in digital domain. Google returns no specific information on this I have read through a lot of forums.
Basically I know it can be done in digital and I have resorted to that in the past but I would preferably like to keep plugins out of the equation because I am not a fan even of the best ones.
I considered separating the sidechain paths and implementing some kind of look ahead that will predetermine the nasty digital peaks.
And when it comes to mastering - I am a big fan of Don Barltey in Sydney who has been doing the do since the 60s I believe. I have had a lot of recordings mastered by him and the way he deals with this problem is exactly how I have described in the first post - compress and put to tape and then transfer to digital. 99% of his process is through the analogue domain And thats what I like.
My favorite mastering engineers in UK do it exactly the same way.
 
Yeah, as you mentioned, you can do the sidechain trick to make analogue equipment "look ahead".

But, I guess you need access to the sidechain signal... not always easy... a lot of work...

It sounds like printing to tape sounds like the easiest option...

Or hybrid system... digital and analogue...

Have you tried analogue equipment that has "feed forward" sidechains?
 
You need a good HF limiter or deesser. The Maselec MDS2 is impossible to beat. The Derressers are pretty good but not in the same league. Around here there are the DBX900 boards but I don't think those are very well suited for stereo program. For vintage, an Ortofon STL631 or the later solid state STL732. Or a Neumann HK66 or BSB74.

Edit: There is also a BSS 402 that I've seen used for this purpose. I don't think they are made anymore but you can easily find them used at a reasonable price. I've never used one but the deessing options look good.
 
Gold said:
You need a good HF limiter or deesser. The Maselec MDS2 is impossible to beat.

Paul is right on as ever.  Out of control hi frequencies are very common these days.  I agree with Paul that the Maselec is a clas leader but I have links with the company so take my opinion with a grain of salt. 

It's also possible to use a plug in, generally on the source before any other process works best for me.  Spitfish is hugely popular and free but I'm one of the few who doesn't like it.  Some people enjoy the Sonnox, again I was underwhelmed.  When I need to do it in the box I activate just the top band of the Waves LinMB, it works well but the Maselec beats it.

Good luck,
Ruairi
 
Anyone tried the Kush/UBK modded Fatso for something like this (specifically its "Warmth" section)?  Just curious how flexible a unit like that might be in addition to its compression and saturation sections...

PS - You might also want to bring this up over at "http://repforums.prosoundweb.com/index.php" in their Mastering Forum...

8)
 
Randyman... said:
Anyone tried the Kush/UBK modded Fatso for something like this (specifically its "Warmth" section)?  Just curious how flexible a unit like that might be in addition to its compression and saturation sections...

I haven't used the UBK Fatso but in my experience you don't need warmth/fuzz/distortion, you just need to eliminate the peaky hi frequency activity that is so common in these days of Chinese mics, dodgy converters, "warm" pres and too many plugs.

 
I believe the Fatso's "Warmth" section is some form of dynamic EQ (like a de-esser, but maybe gentler?).  Just curious if the UBK Fatso was any good for such a task along with its other sections (which are aimed at saturation and compression/pumping - or "analog smoothing" to a degree).

While I'm not using chinese mics or dodgy converters, I do have issues with this on my end as well (and my ears are trashed - certainly doen't help my situation! :( )

8)
 
tardishead said:
uh yeh thanks for that. Actually I am looking for a discussion on whether taming these kind of digital peaks can be achieved through analogue - or whether that is just the nature of DAW recordings that they contain these nasty peaks and they have to be dealt with in digital domain.
I would say it's the nature of digital recording, where you don't have the smearing effect of magnetic tape. A cymbal hit is more accurately transcribed today than ever before -not implying any esthetics here. The induced effect is that the crest factor of most signals is increased, and OTOH clients want a crest factor that is crushed to bone. The adequate answer is either the use of very expensive analog gear, or digital technology, which is much less expensive in plug-in form than in hardware.
Google returns no specific information on this I have read through a lot of forums.
Basically I know it can be done in digital and I have resorted to that in the past but I would preferably like to keep plugins out of the equation because I am not a fan even of the best ones.
I considered separating the sidechain paths and implementing some kind of look ahead that will predetermine the nasty digital peaks.
And when it comes to mastering - I am a big fan of Don Barltey in Sydney who has been doing the do since the 60s I believe. I have had a lot of recordings mastered by him and the way he deals with this problem is exactly how I have described in the first post - compress and put to tape and then transfer to digital. 99% of his process is through the analogue domain And thats what I like.
My favorite mastering engineers in UK do it exactly the same way.
I would describe it as "the use of very expensive analog gear", although you may already have a lot of it.
 
I can't wrap my brain around this concept of 'eating transients' with certain analog gear (tape, certain preamps).
Is it different than rolling off high frequencies? How?
Is it better understood as a frequency dependent compressor?
 
abbey road d enfer said:
The adequate answer is either the use of very expensive analog gear, or digital technology, which is much less expensive in plug-in form than in hardware.

The Weiss DS-1 is the only other currently made device that is as good as the Maselec. It's digital and it's expensive. Plug in deessers have been less than satisfactory IME.
 
dmp said:
I can't wrap my brain around this concept of 'eating transients' with certain analog gear (tape, certain preamps).
Is it different than rolling off high frequencies? How?
Tape and transformers have a transient response that smears the harmonics differently. A sharp attack (Dirac style) ends up as a pulse of larger width and smaller amplitude, but still contains all the initial spectrum (within the limits of their frequency response), so the crest factor is much less than the incoming sound. Within the constraints of a given maximum amplitude, this opens the possibility of making the signal louder without increasing the peak amplitude.
Is it better understood as a frequency dependent compressor?
Not at all; the compressor, being amplitude-based, modifies the crest factor independantly of frequency. I'm giving a simplified explanation, one could object that a compressor operates on the amplitude/time binomial, so is somewhat frequency-dependant.
 
Tape and transformers have a transient response that smears the harmonics differently. A sharp attack (Dirac style) ends up as a pulse of larger width and smaller amplitude, but still contains all the initial spectrum (within the limits of their frequency response), so the crest factor is much less than the incoming sound. Within the constraints of a given maximum amplitude, this opens the possibility of making the signal louder without increasing the peak amplitude.

Thanks for a great answer!
Now it makes more sense that different preamps with identical swept frequency response can sound very different.
 
Keeping it analogue why not try Anamod ATS1? With some settings and tape emulation i think you could deal with it.
Or some Cranesong / Dave Hill limiter design.
 
These big peaks can be tamed quite nicely when mixing down to analog.

I have an ATR102 here, currently setup with GP9.

I'd be happy to run your mix to tape so that you can hear what a good analog deck
can do to your mix.

Regards,
Mark
 
Hi Tardishead,

I recently timed a square wave through a simple triode and it looked like it delayed it 20uS.  Maybe you could run the track through more tubes than the sidechain to delay it enough?

Trying to stay analogue here.

best
DaveP
 
I 2nd the Ampex 102.  I use it to  lower the digital hi freq spikes on vocals.  I'm using an old 249 neuman not a china made microphone, and on some voices it can be brutal after you master limit the mix.  102 ampex is also nice on the low end for bass. 

There are dynamic plug-ins by Brainworx you may want to check out  I have not tested their dynamic Eq but it might be able to help here.  Limiting high freq bandwith with a transformer can help with the sound also as someone else mentioned.  And a high end A to D like prism seems to help me with these problems.  Better sound field   

Ribbon mics can also help with the spikes.  Coles and RCA 77ish mics.  These would be for tracking.

You mentioned hearing problems. "ears are trashed"  Have you had them checked to see what problems  your having?
 
To touch on the specific issue of the peak problem.  Have you looked into the SPL transient designer or the TRANSPRESSOR?  I would think you could use one of those to pull the transients down, no?  I know they work great on bringing them OUT, but i think just turning the knob the other way works opposite, and you're still in outboard gear land.
 
If you have recorded digitally but want to master in analogue my logical answer would be to get rid of these HF spikes in the mix while its still in the box, before you hit your DAC.
As I'm a dance music producer I dont do much actual "live recording" with mic's or drums, so I havent really come across this HF spike effect of Digital, but assuming its a result of actually making a recording as opposed to usings synthetic sounds created entirely in the box, then are the HF spikes confined to individual channels, groups or the effect of summing digitally?
Perhaps you could solo the channels and see which individual channels or combinations of channels are most suceptable to these HF Spikes, then put an insert De-esser, Sonnox trans mod, Dynamic EQ or multiband compressor on the culprits to tame them before they hit the master fader. I think this would have less detrimental effect on the mix sound than just trying to adress the issue on the finished mix down.

As I say I dont entcounter this much in my way of working but it would make sense that certain types of source  or recording chain combinations might be more suceptable to HF Spikes when hitting the digital domain than others.
 
Back
Top