Singer/Songwriter Vocals/Guitar and Video

Recently, my band leader Johanna Herdtfelder asked me to do some solo recording with her, just her songs, vocals and a guitar. I agreed and suggested to try using a DSLR camera for video so we would have something on YouTube. Johanna dropped by my living room/Ardour-based studio we recorded her song “Regenbaum”:

Once we had a reasonably good take, we had the idea of me playing a bit of rainmaker, so this is what I did.

Feel free to comment.

All recording and mixing was done in Ardour. The video editing was done in Kdenlive, in which I added the final audio track exported from Ardour.

For the technically inclined: Mics were an Oktava MK012 pointed to the sound hole and a Rode NT5 shooting at the neck of the guitar. The vocal mic is a t.bone sc450 which looks great… FX include Voxengo Voxformer and Voxengo Elephant compressors (Windows VSTs on Linux, don’t do this at home kids… even though it is very much worth it). Also some EQ10Q, of course, and a bit of Calf Vintage Delay. Reverbs are Yamaha REV-X from the 01V96i console which I also use as an “audio interface” hooked up via ADAT.

Naturally there’s quite some crosstalk between the mics and next time I’ll most likely use one of my MK012s for the vocals, too. They don’t look as professional, but they sound great for the price and hopefully pick up less guitar. The sc450 is not very directed.

The camera was a consumer model by Canon, an EOS1100D.

All the best
DrNI

Nice work,

though id suggest using a different EQ i seem to find that eq10q sometimes sounds like its causing phase issues much like when pulling out to much on a graphic equaliser. I dont get the same “phase” issues when using calfs parametric equaliser.

Im not sure if there is actually any phase issues with it, but to me it just sounds like it compared to using calfs eq when doing more agressive cuts.

Its always going to be hard minimising sound bleed when micing up in that setup as you never get unwanted sounds into the cancelation areas of the mic pickup. hypercardiod pattern may work a little better for more isolation,

But anyway sometimes a little bleed makes things better, completed isolated sounds can become somewhat disconnected from the performance. If you want a really well polished and produced tight sound then yes but i think a more open setup suits acoustic performances.

The mics arnt far enough part to cause any real phase issues i wouldnt think not like when using overheads at a distance for drums where you might have to delay them if you find it partially cancelling out the snare drum.

With EQ10Q there’s the issue (at least in my version) that it can cause pop sounds when input changes from digital black to actual signal, but apparently only in combination with a high pass filter. For future work I’ll do some testing with the linuxDSP plugins. My strategy is to keep the production environment stable and to know my tools quite well. There’s one other thing I’m currently working on, and once this will be finished, I’ll update my Ardour version and check out new tool such as linuxDSP.

And of course there was a phase issue in this recording. Make sure always to use a correlation meter when recording one source with several mics! However, I’m not sure if the phase issue was due to the mic positioning or due to the fact that those two different condenser mics internally also have different preamps.

Btw, for this particular recording, I used IQ4GUI, a dynamic equalizer that unfortunately is quite unstable as Windows VST (the Voxengo stuff is more or less solid). I didn’t find any dynamic EQ for native Linux. It is not so seldom that you have annyoing screaming frequencies in female voices once the ladies get loud. :sunglasses: So dynamic EQs will solve the issue it in that case. Some say, they also use it as a de-esser.

Hi,

Great work! Very nice to see a production like this done with Linux. As far as phasing the YouTube compression always makes it much more apparent even if it is barely detectable in your original Ardour export…

A beautiful singer with a beautiful voice … who cares what the mix sounds like ? :slight_smile: It sounds great, thanks for posting it!

(blatant plug) : If you need a dynamic EQ, Harrison Mixbus has 2 plugins ( XT-MC Multiband Compressor and XT-DS DeEsser ) that you might find useful. They are LV2 plugins and may be used in Ardour. Currently they must be manually installed by first installing Mixbus and then dragging the plugins to your LV2 folder. If you use AVLinux, this is already done for you.
http://harrisonconsoles.com/site/xtds.html

-Ben

Ben, thanks for the advertisement. So this basically means, if I buy MixBus, I can use its LV2 plugins in Ardour? Is it also possible to simply buy the plugins without MixBus?

For me and myself, I bought the Voxengo stuff, which runs almost solid as WindowsVST. But I’d really like to advise other people to buy native plugins. (Voxengo Elephant is a multiband compressor, Voxengo Voxformer is eq/compressor/de-esser that is designed for vocals, but also works well on other instruments.)

Check out IQ4gui. It might not be the best sounding dynamic EQ but its GUI concept is really good. Very easy to use and very clear to see when gain reduction drops in etc. - I’d really like to have something like this in perfect as LV2. :slight_smile:

De-essers and multiband compressors are related but not quite the same.

Oh and btw beauty: I think beautiful voices deserve a sound guy who does the best he can do. I’m still learning… ever will be :slight_smile:

-DrNI

I’d also be interested in using mixbus plugins in Ardour. A standalone version in LV2 would be wonderful.

On the subject of dynamic EQ, the LinuxDSP MBC2B is a multiband compressor that I found works well for getting rid of some of the harshness in vocal performances (It’s stereo only, but ardour doesn’t mind that). I find the gui a but cumbersome, but it’s no trouble to use. I’ve never really got on well with Deessers, fortunately working on computers these days makes it easy to get rid of the esses with region gain, I don’t normally bother with frequency dependent stuff for esses unless I have a really badly behaved singer!

For EQ in general LinuxDSPs black EQ is pretty good and this one has a very nice gui :slight_smile:

Oh and I agree, beautiful perfomance, nicely captured!

@DrNI and @bdp

You can buy some of the mixbus plugins seperate from Mixbus itself, the ones Ben mentioned this is true of. However I am not certain if there is any ways to install them other than via installing Mixbus, which requires a purchase at this time to my knowledge, unless you are on a particular distro such as AVLinux. These plugins should be usable in any LV2 compatible host is my understanding, but I haven’t tried it myself.

         Seablade

Thanks, I had a look at the site. I may give some of the plugins a try when I get some time to play.

Does anyone know if the native Mixbus effects such as the Tape saturation and channel EQ are implemented as LV2 and so could be used in Ardour by purchasing and installing Mixbus? I used Mixbus at one of our local studios which offered it as an option. I really liked it but just prefer the workflow on Ardour 3, but I can see that it would be nice to use the built in effects as plugins on Ardour if possible.

@bdp

No those are implemented in Mixbus in a way that cannot be done easily or effectively as a plugin is my understanding. May be worth looking at MBv3 when it comes out as that is based off of A3 instead of A2.

       Seablade