Fixing Clicks in Sample Editor

Hi guys,

I know I hijacked someone else’s topic on this issue (which maybe wasn’t nice), but I wanted to see if I could get any fresh info with a new thread. I have been recording my piano with much success, except for occasional pops/clicks in the waveform. I am recording in CE 5 with my CI1 and two mics: one condenser and the other dynamic. The click is on both channels and not at a punch in/punch out mark. My question: if it IS possible to fix this in the sample editor, where would I find the click in the waveform? There doesn’t seem to be anything out of place once I have zoomed in far enough to redraw it with the pencil. By scrubbing I have been able to narrow the place down quite a bit. Does anyone have experience in this area? It’s frustrating to have nearly perfect recordings and then hear annoying clicks that ruin it----even in the mixdown!

Thanks!
Bane

Does it only happen when you record? Does it also happen during playback of an VST instrument?

Hi Elektrobolt!

Yes, I’m afraid that this is only with recorded audio. As for the source of the click, I suspect that it’s some kind of soundcard related issue, rather than a mic spike or a CI dropout. Here are some screenshots that show one of the suspected areas of interuption. I think I counted four or so in the entire two minute song.

Zoomed Out a bit:
Latest Waveform.jpg
Zoomed Some:
Zoom Some.jpg
Zoomed Closer:
Zoom more.jpg
If there’s any other info that might help, please ask!
Thanks again!

I know, nowadays people listen with their eyes… But how about posting the audio file…!?

Then we can look with our ears :laughing:

On the most zoomed in wave you posted it suspiciously looks like the tail end of a click! the little sharp decaying component on the bottom of the waveform, increasing in amplitude as you look left. If so then the click would seem to be of some duration and has s starting point earlier than the posted display.

The main question should be “where is this click coming from?” as there are a number of point that clicks can occur, starting at the Mic and working towards the sequencer and power supply.

Check this out:
Chromatic.wma (93.3 KB)

Do the segments where the clicks occur repeat somewhere else in the song? If so, you can cut and paste a fresh part in their places. If not, can you re-record just those segments and paste them in? I don’t think you can edit those out by other technical means without some kind of sound imperfection.

Hi Suprawill1,

There are situations where I could maybe paste other wavefiles to replace the defective ones, but many of these parts with similar notes have different velocities, expression, etc., thus forcing me to grab from another take. I just have to ask, what is this pencil tool for then? I’d also rather not have 15 different audio parts (which it would quickly mount up to with two mics) mixed throughout the track, as it is difficult even when crossfading to get more than a couple of clean takes to mix with the original without clicking themselves at the overlapped area.

As Split suggested earlier it is best for future recordings to find the root of the problem. I have been having CPU spikes occasionally while recording, and my computer doesn’t actually quite meet the system requirements for Cubase or my CI1 as far as my CPU processor, which is a dual-core GHz 1.87.

Has any one been able to successfully correct it on the WMA file above? :slight_smile: (It wouldn’t accept WAV.)

It is always best to find the root problem but in pursuit of your repair questioning, a different take is the way to go in my opinion. You don’t have to retake the whole song, just the section(s) with the clicks. I do this all the time with vocal parts that have unwanted mouth sounds that can’t be edited out. The unwanted sounds can be reshaped or resized but will always retain their makeup.
Do a few 2 bar takes of that section, pick your favorite and then paste it in. (delete the others) You may want to do a larger take to develop momentum/flow.
Make sure you make the cuts at zero crossing of both waveforms and 99 times out of 100, you won’t even need cross fading, even though it’s there for those stubborn cuts. I pretty much do this blemish free on a routine basis.

By listening to your click, I pretty much figured your problem develops during recording from asio spikes due to an incompatibility, latency or lack of system requirements. The latter is what you first need to address to track down the root of your problem. Once that is accomplished, we can further a process of elimination.

I’ve never used the draw pencil in the audio edit mode so I can’t confirm whether or not you can repair from that. My guess is you won’t get it all out from that process.

Cool stuff, Suprawill.

On the take that I am currently working on, I had made a mistake and immediately recorded over the incorrect part. The crossfade made it a smooth transition upon punch in, but I have been working until now to fix the click/gap upon punch out. I did have to use the snap to zero crossing and a crossfade to take that booger out!

It is disappointing that I am going to have to rerecord over any clicks. How would I go about eliminating the system requirements from consideration? I am recording at 24 bit, 48kHz, and 2048 samples for best quality.

My backup plan is to use my Wavelab 7 trial to correct it, either with the Restoration Suite plugins or the Spectrum Editor, although I would rather fix it right here and right now in Cubase. Thanks Suprawill; would any one know what if the pencil could fix it somehow for this take?

Don’t. Those artifacts are too severe for any plugins or automatic corretion tools. Just do what people suggested:

  1. copy-paste from (more or less) identical part
    or
  2. re-record the bar (or 2) around the “click”

No way! You can fix clicks of 1-3 samples (or so) with pencil. Done it many times. But your “click” is spread over thousands of samples.

Okay. I am going to test my system with that Latency Syncon thing listed in the CI forum. I knew I wasn’t seeing any way to correct a click spread over so much time. But really, what’s going to stop the measure from getting a click upon re-recording it? I know, at the rate of one or two per every 30 seconds, it’s highly unlikely, but still annoyingly possible. What steps can I take to find the cause of the click? I would post this kind of thing in the CI forim, but everyone knows I’m most of the help over there. :mrgreen:

As far as punching in and out, I don’t even record over the original track. I use a separate track to record the patch and then cut and paste it into the original track. That way you won’t have to worry about crossfading, punch-in/out gaps etc. Just put the 2 tracks right next to each other and zoom in so you can find that sweet spot between the two tracks that is at zero crossing.

Fault finding is a process of elimination, narrowing things down till you’re left with the fault.

The click you posted actually sound like a series of closely spaced clicks,

You need to eliminate the hardware ( mic, soundcard) first then if they aren’t the problem then computer/software configuration

For example, does the click occur on one mic and not the other, does the ASIO meter register a spike on record at the click, do you get clicks occurring when recording with nothing connected, does it get worse with more that one track in record. This will help eliminate possibilities.

As your computer is slightly underpowered then the software setup is more critical, although it should be capable of recording many tracks without problems. My old PIV 3.2 can record many tracks without problems using C6.

Thanks Split.

I won’t be in a position to try this stuff today, but will get at it asap. A couple questions. When you say: the ASIO meter registering a spike on record at the click, where is this ASIO meter? I am using the tools for CI driver. I confirmed that the click is on both mics simultaneously.

After my last post I decided to find the article on optimizing window for DAWs. My next step will be to update my system drivers and run that syncon test. If the Syncon test returns negative, I will probably get a powered hub for the CI.

ASIO meter is the vst performance meter.

The ASIO meter doesn’t register a spike at the location of the clicks. It stays mostly between 20 and 35. As for recording with no mics, I recorded for two minutes with everything appearing okay except for frequent breaks in the waveform. :confused: Is this normal?
Broken Silence.jpg
I just downloaded the Syncon test for further testing this afternoon. I also plan on doing the tests to conclude if the clicks are more frequent with more mics. Can we determine any problems from the tests I have done so far?

Edit: I know you probably think I’m crazy, but it shows up better in Cubase rather than in the Paint screenshot. If you look closely, you can see the line gets thinner in some spots. I don’t know if this is normal since I don’t do a whole lot of recording silence! :mrgreen:

You will not get the VST activity meter to spike on the clicks during playback. That happened during recording.
Try recording for a while using 2 tracks simultaneously (or more, that should make it worse) as you did when recording the piano. Closely watch the asio meter during that time.

Alright Strophoid, I will test. Since I am using the CI1 I won’t be able to record with more than two inputs at one time. Of course, I could create more tracks and assign them to one of those two outputs, but would that really change anything like what we’re looking for?

No idea, probably best to just stick to 2 tracks in that case.