The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Pops ´n Esses
Old 29th March 2005
  #1
Lives for gear
 
Ruphus's Avatar
 

Pops ´n Esses

I have the Stedman / Royer popfilter lying around since quite a while, but havn´t used it yet, fearing anything between source and capsule ( would almost tend to mount off the mics grill instead ).
So far I preferred to rather watch out for the singer to not blow towards the mic, which all in all has worked well so far.

What I´ve been left with is some occuring esses, especially from sharp female voices. Because of that last week I ordered the best de-esser box that I know of.

While waiting for it I came to think about if a pop filter could had helped with avoiding esses in the past recordings. ...
But from what I imagine a pop filter can´t really do much against esses, unlike with pops, because esses have not much to do with moved air.

Am I right?

Thanx,

Ruphus
Old 29th March 2005
  #2
Lives for gear
 
Darius van H's Avatar
 

I believe esses that are considered annoying are often the result of distorion in the hi-frequencies....that's why you can sometimes de-ess all day long but the distortion (and annoyance) won't go away, only be eq'ed........
Old 29th March 2005
  #3
Lives for gear
 
Ruphus's Avatar
 

I use gear that distorts only little yet in the corresponding range.
What disturbs me is less about distortion, but the HF contents that jut out of the mix.

I don´t know how you conclude that de-essing wouldn´t work. Must be a common procedure to be mis-estimated by all the engineers out there.

Using FFT to cut out the offending contents has worked relatively well here. But also engaging a TC comp for de-essing worked not that bad ( slightly better even I would say ), while a FishFilet plugin named Spitfish didn´t work at all, dulling the whole range.

Anyway, I want something that effects the concerning FQ while leaving the remaining contents untouchted as much as possible.
So I ordered a device that won´t use the common technique through compression, but which works by cancelling out with feeding back the offending range polarity reversed.
I´m expecting close to optimal results with it and might post samples of before and after files.

What about my assumption about popfilters and esses, anybody?

Ruphus
Old 29th March 2005
  #4
JTR
Lives for gear
 

Interesting question

Foam windsocks seem to have more of an effect than just pop filters

Try angling the mic too; that can help, if you find the spot on the pattern response where the highs start to roll off; this will also help with pops too

Mic height is another consideration; getting the capsule out of the singer's direct line of fire; but avoid going too high, or their tone might start to get nasal

Also, this is a bit weird, but some "ess" sounds can come from the gap between the singer's front top teeth; a tiny piece of gum wedged there sometimes will work.

Also, if you are working in a DAW, automating quick fader dips on the esses can be of help - that way you're not actually messing with the sound with eq or compression;
probably a combo of all the above, in small amounts is best bet.

Good luck; what's the fancy de-esser you're getting?

BTW - Darius is correct about sibilence induced distortion; the point is that all that hi freq nrg is very challenging for even good gear to pass clearly ( kinda like the key ring test for mics, you know?) and so what seems like excessive sibilence, on closer scrutiny, is actually subtle distortion components.
Old 29th March 2005
  #5
Lives for gear
 
Ruphus's Avatar
 

Hi JTR,

Thanks for the tip with the gum! hihi Gonna keep that in mind. Always fun to learn new stuff.
Yes, I figured that up from a certain degree any gear would distort and also in the tracks I´m talking of there is some of it, however as I think yet in subtle measures, the actuall noise itself is what disturbs me most.

I´d also estimate that the miking was done near to optimal. Considering the singer the esses captured seem to be along the unavoidable quantum. ( Unless any filter could had helped further ... ) The mic was angled, out of the fireline and slightly below her chin in hight.

( Actually the vox tracks were judged best part of the mix, by an old hand whom I asked for his impression. But it would be great to get it out even better, cause I want the songs for a studio demo. )

The fancy box that I´m awaiting this week is the SPL de-esser http://www.spl-usa.com/DeEsser/in_detail.html, it has been good enough to de-ess Rudi Völler at press conferences and that´s considered quite a demanding application. heh

Best,

Ruphus
Old 29th March 2005
  #6
JTR
Lives for gear
 

BTW; what's your mic & preamp chain?

You know, I swear that with modern mics and digi converters, sibilence is way more of an issue than it used to be; my pet theory is that contemporary B-R-I-T-E toned LDC mics & so-so A-D's are the culprits: Seems to me that darker sounding mics / pres, and the very best possible converters are one solution to this issue.

Well, good luck -

oh yeah, the gum has to be sugarfree, mint flavored Trid --- nah, that joke's getting old
Old 29th March 2005
  #7
Lives for gear
 
bassmac's Avatar
 

What mic are you using? Maybe the mic is just too sibilant for the singer. If you use digital, it can capture more hi end than you may need - don’t be afraid to try some roll off eq as well.

thumbsup

EDIT: JTR seems to have hit send before me, but I think we're both on the same track.

Old 29th March 2005
  #8
Try to record the vocals without compression.

Also on mixdown use less reverb or automate the sends.

Also don't know if you are HP EQing the vocals, but too many people here get carried away with it.

It becomes a defacto standard like using compresion for people.
Old 29th March 2005
  #9
Lives for gear
 
Ruphus's Avatar
 

The chain was a Browner Phantom C through a Manley Slam! into Lavry Blue.
I can well imagine that the gear these days is being more prone to esses. After all we know that the digital world is more accurate ( especially yielding in the highs ).

Seen from there I´m lucky for this vox example that my Slam! drops in the HF range, otherwise I guess there would have rather been some more ess.

Actually am currently speaking to Fletcher about the Slam!s HF loss, but guess he might have missed my e-mail to his office account from last weeks Monday.

Ruphus
Old 29th March 2005
  #10
Quote:
Originally Posted by Ruphus
I can well imagine that the gear these days is being more prone to esses. After all we know that the digital world is more accurate ( especially yielding in the highs ).



Ruphus
Digital at 44.1k is not more accurate than a great analog machine.

Remember their is a filter on the hi's.

Same goes for the bass.

What people hear as more acccurate hi's is the lack of low energy.
Old 29th March 2005
  #11
Lives for gear
 
Ruphus's Avatar
 

Quote:
Originally Posted by thethrillfactor
Try to record the vocals without compression.

Also on mixdown use less reverb or automate the sends.

Also don't know if you are HP EQing the vocals, but too many people here get carried away with it.

It becomes a defacto standard like using compresion for people.
Hi Thrill,

Tracked without compression ( I´m hearing the ELOP to be great especially for female vox, but in my case the signal would only get darker / HF cut dwon one more time ). Reverb was put on only very subtly at mix and seems not to have encreased the problem remarkably. ( Though it became evident that it would if wettened some more.)

No, havn´t HPed during track. The tracks appear well separated and yet there seemed no need for it. Also I like to rather HP in the box if required, thinking that I could shape a curve better to taste that way. ( And since Fletcher once mentioned that HW filters differ in quality I have become a little cautious about it. Not meaning though that I wouldn´t try out HW in case needed.)

Besides, would HP directly effect essing problem, or would it just be meaning that a thinner overall image were to emphasize the essing impression ( subjectively )?

Thanks a bunch,

Ruphus
Old 29th March 2005
  #12
Lives for gear
 
Ruphus's Avatar
 

Quote:
Originally Posted by thethrillfactor
Digital at 44.1k is not more accurate than a great analog machine.

Remember their is a filter on the hi's.

Same goes for the bass.

What people hear as more acccurate hi's is the lack of low energy.
This applies to tracking at 44.1, right?
I track with 96k and would think that the preserved highs were still widely being conserved when sampling down ( in contrast to 44.1 recording ).

Still, I understand well what you are saying about masking. Have experienced it a couple of times. First you think the other signal was brighter, but after listening closer you often find that it is just about less bottom than in the "darker" example.

Ruphus
Old 29th March 2005
  #13
Lives for gear
 
Ruphus's Avatar
 

Quote:
Originally Posted by bassmac
What mic are you using? Maybe the mic is just too sibilant for the singer. If you use digital, it can capture more hi end than you may need - don’t be afraid to try some roll off eq as well.

thumbsup

EDIT: JTR seems to have hit send before me, but I think we're both on the same track.

Hey Bass,

Due to my preferred pre currently losing some highs my mixes at time are being rather too dark. ( Ordered EQ to compensate.)
But before the current rig I used to cut HF all the time. Especially, because used to have grainy convertors.
Thanks for your thoughts!

Ruphus
Old 1st April 2005
  #14
Lives for gear
 
Ruphus's Avatar
 

Coming back to this thread I realize that Darius probably is our Jazzius. Sorry for having been snappish, Jazz.( Looking at it I don´t like the answer I gave. Apologies )

Ok, the SPL box has arrived and me went to what must be the most essed segment of one of the vocal tracks and send it through.
Admittedly at first have been a little bit disappointed on what it can do within a setting that would yet preserve the rest of the audio. But it´s apparently that de-essing is being a limited option in general, provided you don´t want to allow perceivable loss in the adjacent FQ.
And after comparing the result with a de-essing example made through the TC compresssor, jumping back and forth between the two and the original clip the SPL example grew more and more appealing to me.

First the TC clip seemed to have smoothened out best, but it comes in with the thickening of the compression, whereas the SPL example stays light like the original while having noticeably removed some of the annoying hiss.

But listen yourself if interested in the matter. ( samples are 96k / 16 bit )

Besides, Darius, if you are willing to share the time I would be interested to hear what you can do with an EQ instead. I´m sure similar can be achieved, but you have only 1 minute ( which is about what it took to find out the setting on the SPL that would fit to this track ) ... ok, 5 minutes for you.
That would be for about same effect.

If you can achieve even more de-essing without effecting the rest of the bandwidth please take all the time you need. I would really like to hear the difference.

Best,

Ruphus
Attached Files

Original.WAV (1.23 MB, 71 views)

TCcompDeEss.WAV (1.23 MB, 73 views)

SPLDesser.WAV (1.23 MB, 137 views)

Old 1st April 2005
  #15
member no 666
 
Fletcher's Avatar
Quote:
Originally Posted by JTR
Also, if you are working in a DAW, automating quick fader dips on the esses can be of help - that way you're not actually messing with the sound with eq or compression;
probably a combo of all the above, in small amounts is best bet.
If you're working with a DAW and have access to a "no phase shift" EQ plugin that you can automate to turn on and off... find the frequency of the offending "SSS"... and automate the plug in EQ to turn it down very quickly only notching the "SSS" frequency from the original signal... best de-esser in town in my opinion... YMMV.

In terms of recording fewer "esses"... the Phantom "C" is a pretty forward sounding mic in the "SSS" regions so perhaps a different microphone might be more appropriate... also angling the mic back like 5 degrees can often be of assistance with "SSS" issues.

[Side note to Ruphus... I'm on the road and have been having a difficult time getting in contact with the person I need to talk to about your SLAM! thing... I haven't forgotten about getting you an answer].

Peace.
Old 1st April 2005
  #16
Lives for gear
 
Ruphus's Avatar
 

Don´t have such a plugin, but will certainly try that out when I can.
Thanks for the 5 degree vertical angle, will be trying that too for sure.
Have you tried the SPL? ( Listened to the examples? Could it be done much better?)

Thanks a bunch for not having forgotten of the Slam! :O)

Best,

Ruphus

PS: The only LC alternative I have is the U195. She sounded nice through it, but the nostalgic flavor didn´t suit as good to the style and arrangement like the Phantom. ( And maybe I like that mic too much for its clean opulence and possibly am biased. Though I demoed both to the artists before decision and they agreed.)

Maybe I should had tried the SM57 too. Worked great with one of the female voices last time.

PS2: I work in Samplitude. A Behringer class in automation.
Old 2nd April 2005
  #17
Lives for gear
 
matucha's Avatar
Listened to the samples at low volume on monitors (it is too late), and in the headphones.

I don't hear problem in the esses (at this volume it is no surprise - have to turn it up tomorrow).

However I can hear obvious boxyness ... sort of combfiltering/room reflection that makes it to sound weird. There is also some low rumble... Maybe the problem lies here and the combfiltering is the unpleasant thing you hear and try to fix. I'd record it again at different position if you can. Try to avoid reflections from any obvious surfaces like a table or whatever... I'd also placed the singer more near the mic to get more proximity eff low-end, this is tad too light for me and it could help with the combfiltering.


good luck

Matous
Old 2nd April 2005
  #18
Lives for gear
 
Ruphus's Avatar
 

Hey Matous,

You´re right the esses aren´t too prominent at average level. I didn´t even realize during tracking. ( Teached me to be even more cautious next time and not only crank up for control, but also ask the other folks in the room to not making noise for that moment.) But when you crank it a bit and especially as the vox is supposed to stand out of the mix, with the vocal appearing relatively nacked so to say, the esses become much more obvious.

I agree about the boxiness. That is because of the booth. Unfortunately. This while having a great sounding room, but unusable due to environment / leakage in and out. I treated the booth with thick fiber which took away a lot, but a part obviously just can´t be absorbed ( and there remain the glas surfaces from door and window ).

Luckily though the FQs making out the box become almost invisble in the mix. Also havn´t started tweaking the stuff yet. Am waiting for outboard EQs. ( Should arrive early next week from Swiss I hope.) That´s when HP will be done too.

About the other findings I´m heading a different way. Don´t find the vox to be light and it clinges nicely to the dark rest of the arrangement, even yet without having touched EQ.
Same for approximity. I would had much rather loved to put the mic more away. This has been around 50 cm and I would had loved to stretch away for around double the distance. The Phantom works great on distance and for most tasks I´m no big fan of the close miking thing anway. ( Man, I need a good room. Really hope to get that realized.)
Only that putting the mic further would had brought in more booth, that´s why still so close and in the same time more sensitive to esses.

Thanks for the input!

Ruphus
Old 2nd April 2005
  #19
Lives for gear
 
matucha's Avatar
Are these esses really a problem? I just listened to it at normal and high loudness, and I don't hear anything wrong with the esses. Well after some compression etc it could be different.

You have problem with the "SSSH..ame" or with the rest of the esses?

Waves renDeeser works for me in the sidechain mode (without the crossover) in most cases, because mostly I find the frequency content of the esses right, but not the level, so this attentuate the signal a bit at the esssss intensive places ;-) and leaves the freqencies intact.

and I was fooled by a low level yesterday, it isn't on the thin side. Sorry.
Old 2nd April 2005
  #20
Lives for gear
 
Ruphus's Avatar
 

The "shhh" like in "shame" isn´t that strong, but the "ssss" like in "misses the rain" pops out of the whole thing. And like you say I count with it getting more dominant when compressed and cranked up.

I have been and am still trying to work with levels too. But in Samplitude that is a real drag, because its automation is the biggest crap you have ever seen and slicing the audio up to work on object level ain´t that ideal either. Also I´m trying to find most practical ways, because just hate everything related to editing. It takes so much of your time for the least pleasant part of the audio work. That is also why I got me the SPL. Apart of that it works with polarity inversion it also helps you with fixing very fast.

Gotta work with what I have now and I´m still learning. Will be trying out what Fletcher recommended too though provided he has tried the SPL too. As it would be about the same working principle I´d expect it to be similar effect, but shall be asking him about it when we´ll be chatting on the Slam! problem.

Thanks for listening again, Matous.

Best,

Ruphus
Old 2nd April 2005
  #21
Lives for gear
 
Screws's Avatar
 

I cut the offending essses out of the track and put them in their own track, and then put Waves Renn De-esser on that track only. This was Bob Ohlssons suggestion and it works perfectly. No loss of highs on rest of track while esses are managed.
Old 2nd April 2005
  #22
Lives for gear
 
DirkB's Avatar
 

I do the same and sometimes also roll of some top on the s-track...

Greetings,
Dirk
Old 2nd April 2005
  #23
Lives for gear
 
Ruphus's Avatar
 

Very interesting!
Sounds like a pretty good method. That way one supposedly can fx much harder without having to worry about the rest.
Always worth to ask the peeps on GS, so many good ideas out there.

Thanks!

Ruphus
Old 2nd April 2005
  #24
Lives for gear
 
Kyle Ashley's Avatar
 

Ruphus,

I work in Samplitude and I automate at the object level to de-ess. It's pretty easy and quick once you get the hang of it as you can easily spot the sss sounds in the waveform visually. Just zoom in tight and cut the sss into it's own object. At that point, you can lower the volume, eq, and do whatever you need. It works very well, and leaves the rest of the vocal intact. The only downside, is that it can take 15-30 minutes to do depending on the content of the material. You can do the same for other annoying sounds such as p's and b's, t's and ch/sh sounds.
Old 2nd April 2005
  #25
Lives for gear
 
Ruphus's Avatar
 

Hi Kyle,

I´m doing that, but I really dislike to slice tracks just for that. Doesn´t appeal to me somehow. Maybe that is because I´m vaccinated from the past when computer resources would make you trying to omitt too many objects. But it´s not only that. As you know crossfades can help to level out against clicks and pops only to a certain degree. Now, you´d think at least as long as one won´t shift an object from its original position in the track one mustn´t worry too much ... However, it has shown that still even without shifting in timeline you CAN get clicks and pops when the volume levels between adjacent objects and the edited one end up differing too much. That means that you can´t just slice, intensively alter volume and just proceed. You still have to listen concentrated for possible artifacts and eventually go back and play around with start and end of crossfades or adapting crossfade curves.

After all this is an old topic for me which makes me really wishing that Magix crew would finally tid up their host instead of going further in up grades without fixing such incredible mess.

That automation is just plain ****ty. ( I´ve said so often before, I know.)

That anorexic volume line sticking on top of the track and when you want to insert a breakpoint it ends up below -25 db and lower. You have to pull down the whole track line first if you want to insert without altering the entire track. ANd it will STILL be hard to insert without altering in the same time. What a piece of crap!

Or when you want a slanting line for continous change and while you´re at it trying to fiddle that crippled feature suddenly all the objects edges become breakpoints too and there your attempt is gone nirvana.

Why can´t these guys have a look into PT or other sequencers and savvy how it´s to be done?!

If you make a search for "Samplitude" you´ll find a recent and long list that I wrote of what these guys could finally do to remove annoying points like that ( and there are still more than those that I havn´t listed ).
If they did, Samp would become such an incredible host.

But you could just as well ask a cow to drop chicken eggs with that stubborn crew they seem to have there.

Yes, you as a user can work around a lot of bugs or stupid layouts, but why not get it delivered right and spare the time?

At the end they have had enough time to realize weak points like that. And hey it´s not like it would be about a 40$ toy. This is a ( very potential ) software aimed for professional use and just billed for it exactly like that. 1000 Euros ain´t appropriate for a program yet having such imperfect characteristics in basic features.

Granted, they have achieved a couple of really great functions, but now it should be time at the latest to tidying up the mess left over, before working on some new synthesizer or whatever for the ad.

Ruphus
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
Finnish / Music Computers
5
midiguru / Music Computers
14
echotp / Music Computers
2
rynugz007 / Music Computers
6

Forum Jump
Forum Jump