My tracks are getting clipped...help

  • Thread starter Thread starter powderfinger
  • Start date Start date
powderfinger

powderfinger

New member
I record through a mixer onto my hard disk, and when the tracks get too high in volume they get clipped. The same levels going to tape are fine however. Why does it clip on my harddrive but not on tape? The only way I can get it reasonable and not to clip is to jack the volume down while recording and then crack it after it is recorded, but this increases the noise. What can I do? Does my sound card just suck, or can I achieve decent levels in another way.
 
Your setting your levels so they do not exceed 0 dB on your computer right? Recording digital, try and keep your peaks around -1 at the most, sometimes less, depending on the instrument. Tape has the benefit of saturation, you can easily exceed 0 dB on it, without any unpleasant digital type distortion.
 
Not only that, 0db on most tape means that there is several db left until the tape saturation kicks in, and then some until you strt getting distortion. 0db on digital menas that you are clipping.
 
No regebro, that is totally wrong....

:(, oh my, having to clean after you again....

0db in digital full scale DOES NOT mean it is clipping. It only means that the audio has reached maximum volume on those converters and that no further voltage values are possible. The converter has no way of knowing that anything above that voltage value is in fact more voltage or not. You can record audio to a A/D converter that is showing 0dB digital, and still have up to .5dB of headroom available. On the other hand, the incoming voltage may exceed the converters ability to convert it to a value and play it clean, thus, it clips the sound, but that still does not neccesarily mean that you will hear clipping.

I reach 0dB on digital all the time with absolutely no clipping at all. In fact, depending upon if the meters derive it's signal from the actual digital bit stream (very expensive) or analog circuit that is reading the output of the converter (common on most digital equipment because it is cheaper to do, like on most DAT and stand alone CDR burners) you could have audio that shows 0dB on the meters when in fact there is up to .5dB or more of headroom left before full voltage is reached.

Need proof?


Download these three .wav files. They are all mono and 2 secs each, so they are very small.

http://www.echostarstudio.com/audio/tt.wav

http://www.echostarstudio.com/audio/test_tone.wav

http://www.echostarstudio.com/audio/testtone.wav


Now, before you do anything, just play them using media player or something like that, and don't look at the .wav in an editor. While playing them, look at the meters on your soundcards mixer. What does it show?

Another fun thing to do is to output this signal via a AES/EBU out on your soundcard to a DAT player. If you can do that, or even a S/PDIF out, look on your DAT players meters too (or a stand alone CD burner, it is all the same for this test....)

Now, open the .wavs in an editor and take a look.

Tell me what YOUR editor and meters showed between the three files.

One of those .wavs was recorded to 0dB. The other two were reduced by a certain amount from that same .wav, but I will not say how much right now, or which is the original because I want to see if any of you can figure out how much it was reduced with your killer metering and DAW's.

Post your answer regebro.

Next. LISTEN to the files. Which ones are distorting? Can you HEAR a difference between any of the files?

This should at least illustrate that not any of us have meters or DAW's that accurately show what voltage values are actually recorded OR played back in our DAW's.

I was actually going to produce a .wav that had the gain increased by 1dB, but, you can do that yourself and see what happens. You will not hear a difference at all.

0dB in digital means absolutely nothing. If you are tracking and hitting 0dB, that just means that you cannot feed a hotter signal to it, supposedly. But, a few mastering engineers have been actually increasing music to OVER digital 0 for up to 8 samples per second without anyone hearing distortion using the finest D/A converters in the world. So, you can actually go OVER 0dB digital and still not HEAR distortion.

So, your metering does not mean squat in the realm of things. At best, it is a reference for how hot your recordings are. But depending on how good your converters are, and how the meters for your converters actually get and represent the actual level of the sound present in either A/D or D/A conversions can vary widely. Had I recorded this test tone to my DAT player, and just got it to 0db on there, I could have digitally transferred it to my hard drive, opened it in an editor, and more then likely found that I had AT LEAST 1dB more gain possible.

In fact, I opened those three .wav file in two different editors, and one still showed all three at 0dB, while the other showed two of them at a lower value.

Which editor is right?

Get my point?

Ed
 
Powderfinger - while Ed slaps Regebro around the head with a wet fish, I'll have a go and risk another wet fish treatment from Ed. If you're tracking digitally in 24 bits, you've got loads of headroom, and there's not as much need to try to get the levels up around 0 dB. Alternately, if you apply just a taste of compression to what you're tracking, it lets you get those levels up with a safety device applied, and the effect it has on the sound itself is very pleasant, to my ears anyway. I do both - tracking in 24-bit and easing back on the level, as well as applying mild compression, and so it's been a long time since I clipped on anything I've done.

Waiting for the wet fish...
 
I just hate misinformation dobro...

And regebro has been dishing that out from day one. I am not trying to pick on him, just that I see posts like that and I cringe because they just are not true and are very misleading.

Actually dobro, you advice is sound. I am not totally sold on compression while tracking, but will if need be. I ONLY compress to tape if I need to. It is a poor substitute for proper mic placement, and good good, a decent preamp, and an understanding of what actually sounds good!

Download those .wavs and have a look. That is a 100Hz test tone recorded at Digital Full Scale. It plays back at 0db, and there is not distortion that you can hear, thus, the clipping, if any, is not prevelent. But, digital full scale does NOT mean that you clipped the signal, and what you meters tell you is far from the truth in digital in most cases. From converter to converter and software to software, the metering is going to change. It is impossible for you to know in digital if 0dB has been reached or exceeded in an A/D converter unless you can HEAR the clipping, and as I have stated earlier, you can actually have transients that last around 8 sample long that are hotter then 0dB that you will not hear the clipping because the transient is too quick for your ear to catch the distortion from the clipping.

True, 24 bits give a "little" room to not worry about levels as much, but you should still be mindfull to get the hottest recording you can before the onset of clipping in digital. Your audio will just sound better when you do.

powederfinger, what soundcard are you using? If you are using a normal old computer soundcard that came with your machine, well, what do you expect from a $20 soundcard? Of course it is going to sound noisy and not allow a good hot signal. You will find that as you start spending a bit more for A/D/A converters, you sound will improve immensely.

Good luck.

Ed
 
Ed...

Hopefully you have explained a quandry I am now in. I "remastered" a live boot CD using EQ and compression. I cannot hear clipping when I play it through my home stereo but the waveform in Soundforge shows the output above 0dB in certain loud sections even though the majority of the peaks VISUALLY indicate 0dB or less throughout the mix. I've been wondering if I should remix it BECAUSE it APPEARS too hot but I am really not sure...

Is this in any way related to the situation you are describing or as a newbie am I REALLY going to get whacked here??

zip
 
Ummmmmmm...

The rule I always live by is if it sounds good, it sounds good.

If you are satisfied with the sound, don't worry about what Sound Forge has to say about the levels. The meters on my DAT player show 0dB on mastered material I feed it even though a close inspection on a .wav editor show that the files never go above -.1 db! So, I literally have another tenth of a db before it is at digital full scale, even though the meters say otherwise. In fact, I can output a .wav file to the DAT via a digital connection with peaks as low as -1dB down and the DAT players meters will still show 0dB on the Peaks!

Basically, I don't trust most meters. Accurate meters are very expensive, but are the only true way to tell just what the heck is really happening.

On most digital equipment, if you are seeing 0dB, you could very well have at lest .5dB more headroom, then when you consider in the fact that you could exceed 0dB for very short amounts of time without any audible distortion, well, you could assume that you have even more headroom then that.

Use your ears. If you have decent monitors, and you don't hear any distortion from clipping, then there is not clipping. Who cares what the meters or editor says....

Ed
 
Ed,

This may sound really stupid but I'm new to this so here it goes... By meters in SF do you mean the "picture" which comes up on the screen when you modify a .wav file or is there another tool I'm not aware of?

The picture I am referring to seems to go off scale for segments which look like 1/4" in length...I have NO idea how many dB above 0dB that may be...

For an example what I would consider 0dB but the "picture" of the waveform is a peak which just touches the boundry. I have also looked at the meter using the plug-in Maxx Wave Bass and when it plays in bypass mode the meter briefly goes into the red during the loudest passages (bass drun kicks etc) but most of the time is in the yellow sector.

For a reference I downloaded another commercial boot - a show which was radio broadcast fron the same band - and it also went in the red usind the same Maxx Bass plugin.

Is there a better tool I can use for a visual check to be sure. I do understand what you mean about trusting your ear but with a live show with lots of distorted guitar it's difficult to separate guitar distortion from distortion I may have created by having my mix too hot.

Am I babbling here?? Does this make any sense...I haven't had my morning coffee yet...LOL

Thanks. zip
 
c7sus...

No they don't but when I opened the Meters I was able to set it as an option...too bad I'm learning it after I've mixed it...:( Oh well - I really wanted to remix a 3 hour concert anyway....gulp.

Is there any way to set the meters to read positive values?? Anything even slightly above 0dB clips and you have to reset. I'd like to try using sonusmans' reco of trying to push it up 0.5dB or so at least as an experiment. Can the SF (or Cakewalk Meter do that??)

zip
 
Coincidentally........

I've just been scratching my head trying to figure out the best way to increase the overall volume of my song. I don't want to use any more compression, and the peaks (entirely due to the lead vocal track) are reaching 0dB. (I'm using N-track.) Most of the peaks in question just consist of one or a few of those little lines (I hope I'm not being overly technical). I was thinking of going back through the vocal track and "drawing" down the volume at each of the peaks in question. Am I better off just letting it clip a dozen times as long as I don't hear any distortion?
 
Well the Vegas meters will show how much past 0 dB you went down to 0.1 dB resolution.
I've done test mixdowns where I forgot to mute some tracks in the project and ended up with +7 dB. Clip city.
But before I messed with individual peaks the way you've suggested I'd listen to them one at a time. Some peaks that look like that have some real content that you don't want to just throw away. Others don't.
 
Re: No regebro, that is totally wrong....

Originally posted by sonusman:
0db in digital full scale DOES NOT mean it is clipping.
[...]
The converter has no way of knowing that anything above that voltage value is in fact more voltage or not.

And neither have you. So for all practical purposes, you should never reach 0db when recording to a HD, since you most likely are clipping (unless of course you have a software were 0db isn't the maximum, there may be such a software, but the ones I have seen show do have maximum an 0db.)

oh my, having to clean after you again....

You are not cleaning up. You are pointing out things that, although strictly correct, almost always are of no value whatsoever. When you are NOT nagging on me, however, you give a lot of very good and valuable advice, so I have so far let it pass. But this time it sounds like your advice is contrary to mine, which is that you shouldn't try to get 0db levels when recording to an HD. And if that IS your advice, it is bad advice.

Get my point?

Well, If your point is to try and nag about details just to try and prove your superiority instead of helping powderfinger out with his problem, I do get the point and I think it's a rather silly one.
If you point is to give powderfinger the advice to record up to 0db, it's bad point your making.
If your point something else, then no, I don't get it.
 
Thanks Ed and Doc.

Doc, I don't know what you mean by "Vegas meters". The meters in N-track don't go above 0dB, as I recall.
 
Hey Reg,

Although Ed buried a lot more detail around it, I think one of the main points he was making was the very correct one that meters DON'T tell the whole story, but the ears certainly do.... the clip points on some s/w and h/w is not necessarily consistent from manufacturer to manufacturer. Some like to build-in a slight amount of headroom at their 0-point, others take the hard-line and say 0 = distortion. Engineers need to take time to learn their gear and how it responds to signals - experiment by pushing a hot signal out and finding exactly where the limit is - you'd probably be surprised at how much "give" there actually is! When I get a new piece of gear, before I even check out all the cool things, I try and find out exactly how far I can push it, and how low I can go before the noise floor overshadows things - this way, I know exactly what to expect when I throw something at it... it also depends on the type of signal too - high-transient signals need more "breathing space" at hot levels than say bass (unless it's a pop/slap bass line!).

As for the other stuff, can't we 'jes stawp the fussin' 'n a'feudin' 'n all 'jes git alawwnnnggg??? :eek:

:D

Bruce
 
Sorry, not biting here regebro....

0dB DOES NOT mean clipping in ANY digital recording system I have dealt with, and that has been many.

I have never seen a case where the metering was too generous, always, they show the audio to be higher in volume then it actually is, if not exactly what it is.

Bad advice? I spent all that time recording and uploading three .wav files to my website. It was recorded to Digital Full Scale. Where is the clipping in any of those files regebro? Did you even bother to LISTEN regebro? Come on, tell me how bad they are clipping. There is absolutely no possible way I could have recorded them hotter then they are without throwing too much voltage to the A/D converters, but for sure, they reach digital full scale.


Yes, I am nagging you.....:) I will continue to do so if you put out bogus advice. Trying to achieve digital full scale while tracking IS NOT bad advice. I and several other people follow that "bad advice" and the results are documented for all to hear regebro.

Funny, you have 700 posts, dish out advice like you are an expert, yet have not one link to a mp3 of a recording that I have ever seen on this BBS. In all fairness, let's hear the results of you following your own advice, and how much more killer the audio sounds because of it.

Anything more that I say would start implying stuff about you regebro, and I "try" not to do that (although I lose it here and there) so I will keep this civil and stick with the facts. My facts are correct in this case, and in many others, your's are not.

Be careful what you write, you never know who is reading.

Ed
 
I have yet to see one place where you correct my "bogus" advice. I'm trying to help somebody with a problem. You are nitpicking and giving the advice to always make sure you hit the maximum on the recording meter while recording. That is not good advice, since, as you correctly point out, when you are hitting maximum, you do not know if you have hit maxium or are clipping.

Yes, If you KNOW that your equipment have another 0.5db of headroom over the maximum on the meter, then you can, after adjusting so its slightly below 0, push it carefully up just a nudge. Now, does somebody who gets nasty distortion or loads of noise have that kind of intimate knowledge of the equipment? No.
Again you give advice which are either bad, or requires loads of expertise and experience, but you are giving it to a newbie. That isn't helping.

If 0db is not the maximum on most computer recording equipment, but only on the ones I have used, then i apologize for making that assumption out of the experince I have had. However, you did yourself write "0db in digital full scale DOES NOT mean it is clipping. It only means that the audio has reached maximum volume on those converters and that no further voltage values are possible" . Was that wrong? Is 0db in most cases not the maximum of a meter in computer recording? Then what is the most common? +0.5db? +3db? +6db?

Btw:
I have SEVERAL times put out a link to some mp3's here, and if you'd bother to even look at my profile, you'd see the link there. It's recorded on a 4-track in my apartment. If you search in the mp3 forum you can get a more detailed equipment list if you want it.
 
Know what you do not know....

Those are words that keep me learning regebro. If you are going to share what you don't know, but put it across like it is something you do know, then have the balls to argue what you don't know with people who DO know it, then be prepared to be corrected, and in sometime harsh and somewhat rude ways. If that is your style of learning (and having been involved with teaching or training in one way or another for many many years, there ARE people who only learn in this manner) then have at it! I will continue to share it the way I know it, and you can, well, do whatever.... And trust me on this regebro, I have been patient in the extreme with your far out "advice" you dole out like you are Bob Clearwater or something.

As to the equipment that you have used that will not allow you to reach digital full scale while recording, I can guarantee you that it is possible to reach 0dB on those meters while recording and not have clipping. Just because YOU can't do it does not mean that it is not possible, only that you have not figured out some problems with your sounds or mix.

Anyway, no more time to argue your unfounded rubbish on this one.

powderfinger, thank you for the nice email you sent. Hopefully, some of those things I told you to check for will solve the problem. In addition, look for ANY low end instrument to be the culprit. Because of the nature of near field monitoring, many people put way too much low end in their mixes, thus eating up valuable headroom while mixing, thus, lower appearent loudness in the mix, etc....

Good luck.

Ed
 
You must log in or register to reply here.
Back
Top