If you’ve read my posts about mic comparison you know what I mean when I say comparison – all the mics recording the same source in as close to the same location as possible, with the level of the recorded tracks matched as carefully as my equipment and patience allow. Here’s the mic array I set up for the comparison:

Four mics ready to test

The mics I’m comparing are the two I mentioned before, the Rode NT1a and ADK A6, plus a couple from my collection. I chose a Schoeps CMC54 as a kind of reference standard, and for something a little different, an ElectroVoice RE15 dynamic.

I used AudioTools from Studio Six to generate a 1 Khz test tone. All the mics were connected to the preamps of my RME UFX interface. The wonderful TotalMix software let me fine tune the preamp gain to get a pretty close level match. Then I recorded the test tone simultaneously on all four of the mics and used the Sonalksis FreeG trim/meter tool to fine tune the levels. With the tracks trimmed I then recorded a bit of slack key using my Martin OM18GE and rendered each track as a separate 44.1/16 WAV file.

Hear the Difference

As usual, my comparison clips are uncompressed WAV files. They won’t stream very well if at all, so you should download the files to your local system before attempting to compare them. It’s also really valuable to listen to these in a double-blind comparison tool like Foobar2000 and its ABX comparator. Not only does this tool hide the labels, it also makes it easy to concentrate on small sections, hear them repeatedly back to back, and track the statistics of your choices.

r.wav
s.wav
t.wav
u.wav

So how do these sound to you? Do you hear one of these mics as noticeably better than the others? Notably worse? Please leave a comment here if you’d like the key to the comparison.

Then I ran across an Ebay auction for a Martin Tiple. Now the “tiple” is one of a number of smallish stringed instruments played throughout the Spanish and Portuguese speaking world, where the word is pronounced “teeplay.” But a Martin Tiple is a strange member of the `ukulele family, an instrument that was vitally important to the CF Martin company in their history. I’ve learned a lot about the Martin Tiple by visiting Uncle Emile’s Blog, The Martin Tiple.

I started off tuning this tiple like a modern tenor uke, G C E A or C6 tuning. But we discovered that it plays and tunes better one step higher in a D6 tuning of A D F# B.

Beauty Shots

There’s something about the organic glow of a fine vintage instrument, the battle scars contribute to the feeling that this is a special object, simply by having survived. The instrument I received from my Ebay seller has been loved pretty hard and had a fair amount of repair work along the way, but the cost reflected that so I’m happy with the result. Martin built these instruments in their common styles of 15/17 (mahogany top, back and sides), 18 (spruce top, mahogany back and sides), and 28 (spruce top, rosewood back and sides). I’ve seen tiples as early as 1925 and as late as the 1970s, this one is from 1949 according to the serial number database on the CF Martin site. It’s easy to see why people fall under the spell of old Martins.

Old spruce, old mahogany, a bit of old rosewood.

The peghead is one of the most distinctive in the world of stringed instruments, with those five-on-a-plate tuners marching down each side and the forest of strings scurrying over the nut.

The tuners are recent replacements. All the originals I’ve seen had white buttons.

This instrument got plenty of legitimate play but not abuse. Looks like someone really had fun with this tiple. If you look carefully you can see that the saddle has been compensated slightly for each course, but not for the intonation difference between the plain and wound strings. This makes tuning accuracy a bit marginal.

Someone enjoyed strumming this little baby.

The bridge is a block with wings, carved from rosewood with holes drilled through for string attachment. The angled bone saddle drops into a slot in the rosewood.

This saddle is about as low as it can go.

The ball ends serve to hold the string in place against the bridge. This arrangement gives a very shallow break angle over the saddle, probably adding a bit to the jangle and chime.

The string fastening system is a simple hole through the rosewood bridge.

There are faintly visible lateral waves in the mahogany back, I can’t tell if it’s a figure in the wood or saw marks from the resawing, but in any case it’s lovely to look at.

I can’t tell if the wave is flame figure or saw marks.

So What Does a Tiple Sound Like?

The sound of this little box is pretty amazing. At time it sounds like a 12 string guitar, sometimes like a mandolin, or a Tahitian banjo. It really came to life when Hawaiian slack key and `ukulele master Ledward Kaapana showed up at our house. Ledward told me that one of his Aunties back in Kalapana featured a tiple as her musical contribution to the legendary house parties of his youth. He quickly got the tuning sweetened up and began finding his voice on the tiple.

We’ve been staying pretty busy playing shows around Northern California, but with a few days off I persuaded Ledward to join me in the Digital Duck Studio to do a little video featuring his tiple stylings. I pulled out my Wingert Model E, tuned to taropatch slack key, and we came up with a couple of old Hawaiian songs to share.

How’s that for some jangly fun? If you search around YouTube you can hear some other examples of the tiple being played in a variety of styles.

Recording Details

Since this is a techie blog I should take a moment to address the recording process. I’ve been experimenting with the ORTF mic array lately. It gives a “big” but natural representation of a solo guitar, and I really liked the result in a recent duet recording. So I left the pair of Schoeps cardioid mics in their ORTF arrangement, about two feet in front of our instruments. I used the Panasonic GH2 along with two Sanyo Xacti HD2000 cameras to provide several camera angles.

The audio came through the RME UFX interface and was captured in REAPER. I used no EQ or reverb, the only adjustment was a level boost along with a brickwall limiter to prevent overs. I assembled the audio with the multiple cameras in Edius Neo Booster and rendered as a 720p AVCHD video for upload to YouTube.

A commenter on one of my Zoom Q3HD blog posts asked about freeware tools for raising the level of his video soundtracks. I touched on this some years ago when the subject was syncing audio and video, but modifying the original audio track makes things even easier. I found two free tools that can be used together to enhance the audio without impacting the video of the Q3HD. Most approaches involve transcoding, or decompressing and recompressing the video and that approach risks damaging the video and adds production time to the workflow.

Audacity

Audacity has been the standard free audio tool for many years. The developers have continued to improve this classic and these days the program handles VST plugins and offers guided help in installing the LAME and FFmpeglibraries that extend Audacity’s native capabilities. If you’re following along, be sure to download and install both these extensions along with your favorite VST plugins.

With FFmpeg installed it’s easy to open the audio stream of many video container files, such MOV, AVI, or MP4. Simple choose File | Import | Audio to navigate to the desired file, or drag and drop the file using your OS file manager. Once the file is open in Audacity, you can adjust EQ, dynamic range, and level. You can also add effects like reverb, echo, or chorus. What you should not do is change the length of the file. No trimming of heads and tails, not adding on at the beginning or end either. This procedure is simple because we’re using the original audio file, so we can be confident that it will match back up with the original video, but this only works if we don’t mess it up.

Trim the Spikes

One trick for raising level involves lowering a few large spikes in the audio waveform before adjusting the file as a whole. These spikes are clearly visible in the waveform and usually represent handling noise or other glitches. We don’t lose anything by making those noises lower in level, and it makes the procedure for raising level easier.

By the way, you can open two copies of the audio in Audacity, adjust one of them, and then go back and forth to compare your adjustments to the original.

When you’ve made your adjustments, simply File | Export then choose your output format. I recommend WAV (PCM) uncompressed audio because this retains maximum fidelity.

Avidemux

When I began playing with HD digital video, Avidemux caught my eye. If offers very flexible handling of the video and audio streams that make up the container files like .MOV, .MP4, and .AVI. We’ll take advantage of that flexibility to reunite our original video to our new enhanced audio that we created in Audacity.

Simply open the original .MOV file. Then use Audio | Select Track | Track 1 | *dropdown* | … Add audio track to pull in our enhanced audio file. In my initial experiments with output formats offered in Avidemux I didn’t run across one that includes PCM audio and the H.264 video codec, so in my demo video I converted to a compressed format. We can do it in the Audio | Select Track menu, change copy to AAC then use Configure to set the bitrate up to 224. This should give a pretty transparent conversion, but if we use AVI or MKV for our container we can stay with the PCM audio by using copy. Finally, set the Output Format. In my demo video I chose MP4 because I’ve had good success with it in the past. Use File | Save to select the destination for the new output file, and be sure to set the extension on the file name to match the container format, in this case MP4.

Let’s Look at the Video

I was quite pleased with how easy this procedure is and how effective. Here’s the original video:

Here’s the altered one:
 

and here’s a “How To” video demonstrating the process:
 

Both these tools can do a lot more than we demonstrated here. Audacity, as mentioned, can add all manner of audio effects and alterations. Avidemux has a nice range of filters for resizing, zooming, coloring, fading, and more. These filters won’t work without transcoding, but it is possible to trim video at certain points (i frames) while simply copying the stream. In the past I’ve had stability problems with Avidemux, but this time out it worked without a hiccup. It still doesn’t play H.264 HD video without stuttering, but it’s so handy I don’t mind the poor real-time performance.

Bigger Must Be Better

It’s common to read recommendations of higher sample rates and longer word length, often stated as if there can be no question that the result is better quality. And manufacturers have been quite willing to go for ever bigger numbers, so that we can now buy converters that sample at 96 khz, 192 khz, even 384 khz. Word length has not seen such a dramatic change, since the lower limit on self-noise is determined by the rattling of the atoms in our conductors, somewhere around 124 dB or 21 bits.

I’m generally pretty skeptical of claims of “high resolution” audio because of the simple fact that human hearing is quite limited – only some young listeners are able to detect sounds slightly above 20 khz, and there has never been evidence that higher frequencies affect our perception of sound. In view of this, microphones almost without exception have steep rolloff above 20 khz. Furthermore, I have never seen objective evidence that people can detect an audible difference between music recorded at 44.1 or at 96 khz. Plenty of anecdote, but no objective evidence.

Can We Have the Clips, Please

As it happens, I have two good quality audio interface boxes at my disposal, and even better, one of them can record without being connected to a computer. This gives me a way to capture a single performance at two different sample rates. So I hooked up my Echo Audiofire Pre8 via Firewire alongside my USB connected RME UFX. The UFX has the wonderful DUREC feature which records to a USB device connected to a front mounted port. So I set out to create a pair of same source level matched clips at 44.1 khz and 96 khz sample rates. I used 24 bit word length for both, because that is the default for the DUREC recording feature and because I prefer to reduce the number of variables in a comparison.

Both these interfaces have mic preamps, but in order to reduce the variables I chose to use a single external preamp to feed the line inputs on the two interfaces. I connected a pair of AT3035 mics to an M-Audio DMP3, then into my handy little Coleman Audio LS3 passive splitter. From the LS3 I connected to the line in on the Pre8 and UFX, fired up the test tone generator on my iPhone, and hit the first bit snag.

The line inputs on the Pre8 are also mic inputs – and turning them to their lowest setting does not match the standard sensitivity of a line level connection. Suddenly the simple step of matching volumes got complicated.

I had hoped that the Echo Firewire Console, their software mixer for the Pre8, might come to my rescue, but the metering was too coarse and didn’t include a numeric readout of the input level. Unlike the RME UFX, the Echo unit has no software control of input levels. As a workaround I opened up REAPER, turned on monitoring, and used the Sonalksis FreeG meter to match the AudioFire levels to those shown in the RME TotalMix software mixer. But that wasn’t the end of my troubles. I was going nuts trying to get a balanced stereo image, due to the impact of room effects on the test tone and the variability of my mic position. I tried for a couple of days, but finally just gave up and went to mono.

NOW Can We Have the Clips, Please

With the configuring and tweaking done I tuned up the Wingert E to taropatch open G (D G D G B D) as usual, and practiced a bit of E Lili`u E, a song of praise for Queen Liliu`okalani, last reigning monarch of the Kingdom of Hawai`i. I set the Echo AudioFire Pre8 and REAPER to record at 96 khz and 24 bits. I configured the RME UFX to record at 44.1 khz and 24 bits, with the output to be captured on a USB flash drive. I recorded a bit of 1 khz test tone and a few passes through a verse of my song. Then pulled the UFX file into the REAPER session where the Pre8 96 khz clip had been captured. I tweaked the levels to be within .1 dB, selected my favorite take, and rendered a 96 khz 24 bit WAV for each source. Note that both these files are now 96 khz. I did this in order to keep the test blind, because the sample rate is basically a label, and a label on a sample will inevitably affect what we hear. Since the sample was originally recorded at 44.1 khz, it can never have more information than at that initial capture, the current 96 khz format essentially serves as a container for the 44.1 khz information.

So here they are:

m.wav

n.wav

Please don’t try to play these files by simply clicking on them. The 96 khz 24 bit WAV file is very large for its length and will not stream well at all. To download the file by right click Apple option click and select “Save Link as …”

ABX Troubles – No, Comprehension Troubles

When I originally published this post I was failed to do an ABX comparison of the two clips, because I didn’t understand how to configure my computer to playback the 96 kbps files. I did a quick internet search and thought I discovered an insurmountable problem, so I just gave up. Fortunately a poster at the REAPER forum corrected my misunderstanding, and when I took a second look at the issue I was able to get foobar2000 and its ABX utility working.

I’m so glad I had this learning experience. When I pulled the two clips into the ABX utility I listened through them trying to find a difference I could hear. I never succeeded. But I’m old, my ears have suffered a lot of abuse, and my bias is toward finding transparency, or no difference. So I rely on others, especially those with younger ears, to confirm the audibility of difference or lack. I hope some among you will download the clips and and pull them into an ABX utility and let me know the results. And as always, I’ll send the key by email if you request it.

I recently uploaded a set of clips from our guitar party, encoded as 720p/30 H.264 video and PCM 48khz/16 bit audio. When I pulled one of the 720 streams down from YouTube and examined the specs with Mediainfo I discovered that audio is now returned as a 192 kbps AAC stream. This is generally considered to be the audible equivalent of an uncompressed file for nearly all material. In other words, this level of compression using this codec should sound just like the original uncompressed file.

ABX for the Answer

I still strongly believe that any audio comparison should start with an ABX session, to confirm that differences are really being heard. I have observed in myself and others that even blinded, we quickly attach characteristics to a label, and once this happens we “hear” the label as much as the clip, or more.

Here are two ABX tools, one for PC and one for Mac.

In order to provide an accurate comparison between the uploaded and downloaded versions, I opened both clips in REAPER. I created a very brief extract of each file along with the full file, each rendered as an uncompressed 44.1/16 WAV file. Be sure to download these clips to listen to them, since they’re .WAV they are not intended for streaming.

joe-extract
blo-extract

joe-complete
blo-complete

So can you do 16 comparisons between these files in a double blind comparison and identify them correctly 13 or more times out of 16? Please let me know in the comments and I’ll get back to you with the ID of the clips.

Some Outside Sources

Google/YouTube has provided a bit of updated and expanded information on file formats, with specific instructions for some common video tools, recommended formats and specs, and so on. Check out the Video encoding page.

A gentleman named Nick Vogt blogs about various tech topics, and has a nice set of charts showing the progress of YouTube audio on this page. His work shows consistent progress from the bad old days of gritty lofi to today’s commercial release quality.

I was expecting three or four people to attend, but to my surprise we had guests from Oregon and from the LA area as well as nearly a dozen local players. When we lined the attending instruments up at the end of the event it looked like this:

More than one guitar per person

That was the last time we captured a picture of all the guitars present. Even at this first gathering, the music and interaction with people was much more important than trying out new guitars, and at following events the number of instruments became overwhelming.

We Started a Tradition

Our second party included some recordings, the only one to do so until this year. The second party was also the one that led us to a bi-annual schedule. The Healdsburg Guitar Festival was well established by then, and our luthier buddies were busy with their show guitars, so we began holding events in even years, leaving the odd years for the HGF.

In 2005 we made an exception to our alternate years rule when a regular on the Usenet news group RMMGA (rec.music.makers.guitar.acoustic) planned a rare visit to California. Several folks asked us to host a get together to coincide with her trip and we had a terrific off-year event.

Our party in 2006 was the first with a strong Hawaiian music presence. Members of the halau where Lynn was dancing brought `ukulele

and hula:

Our very special guests were the Brown Ohana from Maui. Kevin and Ikaika Brown were touring Northern California and I had been their road manager on the tour. The became the center of a spectacular kanikapila in our living room.

The parties have continued over the years with a number of luthiers attending each event and a wide variety of musicians and instruments. In a good year there will be four or five jam sessions in various styles all around the house and yard, luthiers trying instruments by other builders and swapping tips and stories. Several rooms will have players sampling instruments they may have only dreamed about before. We’ve seen just about every kind of acoustic guitar, from fine old vintage instruments to the most imaginative of handmade modern guitars, resonators, harp guitars, baritones, classicals, even `ukulele.

Flangdang

One thing that has bugged me along the way is coming up with a good title for each year’s party. Until now I’ve tried to think up something new each time, and this year I was going to use Guitar Fandango 2012 in tribute to the wonderful Spanish Fandango parlor guitar piece. But as I was composing the invitation I remembered that Elizabeth “Libba” Cotten called her version of that tune Spanish Flangdang and I knew that would be the perfect choice. So I’ve adopted the name for future events.

Our Guitar Flangdang 2012 will hopefully be just the first of a new tradition, the Flangdang Video Booth. Although it was a small sample of the instruments and talents present, we captured a number of wonderful performances on video. I uploaded the clips to the Homebrewed Music YouTube channel and created a playlist.

For the Technically Minded

My video booth recording setup was as simple as I could make it. I set up a pair of mics, Audio Technica AT3035s in an ORTF-ish array, into an M-Audio DMP3 preamp. This provided phantom power and good clean gain. I created a chain of adapters to convert the output of the DMP3 into a single 3.5mm (1/8″) TRS (tip-ring-sleeve) stereo plug. Then I plugged that into the line in jack of my Zoom Q3HD. I adjusted the gain on the DMP3 to avoid overloading the Q3HD audio input. With an AC adapter on the Q3HD I could leave this arrangement setup all day and a video session could be controlled by a single button, the red “shoot” button on the Q3HD. I recorded at 720p/30 and PCM (uncompressed) audio on the Zoom.

After the party I dragged the files from SDHC card to the hard drive on my desktop machine. Our 20 or so clips had only used bout a quarter of the 16 GB card. I edited the clips in Edius Neo, adding titles (misspelled ones in many cases), trimming heads and tails with a bit of fade. Since I’d adjusted the preamp gain conservatively I raised the clip audio volume in Edius to compensate. There’s no reverb, no EQ, so you’re hearing the player and the instrument in the room. I rendered using the Edius Quicktime exporter, because that format gave me an option to used PCM audio for my upload. I exported the video as H.264 at 720/30.

I’m really pleased with this video booth experiment, and I’m hoping to have an improved version for the Guitar Flangdang 2014.

I’m quite sure that I was suffering a severe CAS attack when I snagged a GoPro HD Hero2 Motorsports Edition from Overstock.com. I’ve been considering and researching and resisting the GoPro for nearly a year now, but when it popped up on the big O for $249 instead of the usual $299 my will power collapsed. Now that I’m looking around, by the way, the camera seems widely available at this lower price.

So here I am with my new camera and wondering just what I’ll do with it. After all, this a sports camera, an action camera, and my videos are nearly static and so far have been sports free. OK, OK, I should have done more research, should have looked at more clips and read more forums and understood a lot better what one is purchasing when one buys a GoPro.

For one thing, buying a GoPro is just the beginning. The beginning of trips to GoPro accessory sale page. Let’s start by assuming that we’re only interested in the latest and greatest, the HD Hero2. This version has been available since October, 2011 and added 1080 video with variable field of view and a mic input along with other upgrades. This camera is available in three editions, Outdoor, Motorsports, or Surf. There’s no option to buy just the naked camera. Well, actually, the naked camera isn’t very useful, because there’s no way to mount the camera on anything. A housing is needed, and a housing is included with all three of these editions.

Here’s a shot of the goodies that came with my Motorsports Edition:

All the parts included in this Edition

Accessorize

Here comes the first trip to the accessory store, because all three editions come with the waterproof housing. But with this housing the microphone jack is blocked. So to use the external mic input, we need the $39.99 HD Skeleton Housing. And if by any wild chance we would like to use a standard tripod, we might as well buy the $7.99 Tripod Mount while we’re there.

Here are the parts I picked up:

Skeleton case and Tripod Mount

Time to Shoot Yet?

So now I have the bits and pieces needed to shoot a “look at me play guitar” video and I suddenly realize that the fish-eye effect of the wide angle lens is really not what I’m looking for in my videos. And using the narrow field of view option in 1080 results in terribly noisy video. Oooops. I’d like to find a technical discussion of the method used for the medium and narrow angle field of view, but judging by the way the noise and grain jumps up, I’d guess that these modes basically use a smaller area of the sensor, as if the sensor was physically smaller. Using fewer photosites results in less light being captured, resulting in more noise. This jump in noise can be seen in some other modes, like the WVGA 120fps slomo mode.

OK, it looks like the fisheye effect will be part of my videos unless I’m outdoors on a bright day. How’s the audio? Well, with the default sealed waterproof case, the built in mic can barely catch any of the sound of my guitar. Might be OK for engine noises and such, but for music this won’t do. The waterproof case does come with a second rear door, the “open” door, and this does let a bit of sound in, but the result is mono and not too wonderful. A switch to the skeleton case doesn’t improve things much.

One sealed, one open

Plugging in an external mic isn’t a lot better, I’m afraid. There’s no level control, so the audio level will be whatever the camera circuit decides. I got pretty low levels using my Audio Technica AT822 and Rode NT4, both pretty well respected battery powered stereo mics. My little Reactive Sounds DAT mic actually gives slightly better levels, but still less than I would like, and when I raise the volume in post there’s a putt-putt “motorboat” sound in the background of all the external mic recordings. Hmmmm, this is not going well.

So when I set out to shoot I went to the old standby, parallel audio recording. I set up a pair of Audio Technica AT3035 cardiod condensers connected to my RME UFX interface and captured the audio in REAPER. Here’s the result, one of my old favorite Hawaiian songs, Kaimana Hila, as seen through the eyes of the GoPro HD Hero2. I shot in 720/30 mode because that seemed to my eye to give the cleanest image.

In order to show off the capabilities of the camera and have a little fun, I shot the complete song four times. I used the tripod mount twice, for the front and overhead shots. Then I used the Motorsports Edition suction cup mount to attach the camera to the guitar body, and one of the sticky mounts to fasten the camera to the peghead. Then I edited the clips in REAPER to create the final video. The title shot was done by attaching the camera to the Sharpie and shooting a timelapse.

So, I’m honestly not thrilled with the GoPro, and on some level I should have skipped this purchase, but now that I have it I’m determined to have a little fun with it. Maybe fly it on a kite?? Hmmmmm …

Home Sweet Home – Taropatch Tuning

Open G tuning is my home and has been for the last ten years or more. Before I discovered the joy of slack key guitar I thought of tunings as part of bottleneck blues, and I thought tunings meant playing in one key. But a number of Hawaiian musicians showed me that I was mistaken. Kevin Brown of Maui had a huge impact on my understanding of slack key when I saw him play the full range of chords, in whatever key was needed, always in a slack tuning. As I searched and learned I found that the only limit to playing in tunings is the player’s imagination.

Over the years I’ve explored ways to cover different keys, and I got an itch to shoot a little video demonstrating some of these techniques. I wound up with three separate videos, but they should be viewed as a set if possible. Only the first one discusses open G in detail, for instance.

Start in G, Go to C

I’m sort of aiming at folks who are already familiar with open G tuning, but let me start with a chart showing the tuning and the basic I, IV, and V chords.

First position chords in open G

Until recently Leonard’s C, the drop C variation of taropatch G tuning was my main tool for going beyond G. This clever variant goes half-way around the scale with a change to one string, dropping the low D down to an even lower C. Not an open tuning, it requires a modified standard tuning C chord shape to play the C major chord.

Drop C notes and chords

The open G that remains gives us a lot of great ways to play the V chord, though. The rest of our chord knowledge remains useful as well. The scale is slightly altered, with a new emphasis on the F instead of the F#. But having found E7, Am, etc. in taropatch, those same shapes now make the same chords in the key of C.

Here’s a video that lays the groundwork with G tuning, then demonstrates the details of drop C.

Start in G, Go to D

Plenty of harmonicas and accordians are tuned to a specific key but played in a different key – it’s called cross-tuning or cross-note playing. We can use the same idea to play the key of D while staying in our straight taropatch open G tuning.

Chords in D, tuning in G

This particular trick gives us a strong IV chord, G in the key of D, but the V chord, A, becomes a bit trickier. The barre at the 2nd fret works, as do the other shapes that one would use in taropatch, but these either require a barre or lack an easy bass note. This lack of an open A bass limits the kinds of runs that can be easily executed in the cross-tuning concept.

One approach to this problem is to choose a song that gives emphasis to the IV chord. Several of Dennis Kamakahi’s classics feature a lot of I to IV chordal movement, and the cross-tuning concept can really work for “Koke`e” or “Wahine `Ilikea.”

Another trick – rearrange the song. In the accompanying video on cross-tuning I sketched out an arrangement of Hula O Makee that changed a few bars of V chord to the IV chord instead.

Start in G, Go to D Again

The capo is usually the first tool I grab when I need to change key. The simple concept of clamping all the strings to a fret is hard to beat in theory, even if the practice can be tweaky at times. But there’s a way to use a capo that’s just a bit trickier.

I’ve seen articles about cut capos and partial capos for years, but I just skipped over them, because they seemed like a gimmicky way to approach the guitar. Lately I’ve had a change of heart because I’ve found the partial capo to give me a nice sounding key of D while using taropatch open G tuning.

Key of D from open G with a bit of a cheat(er)

The video illustrates this technique, capoing the higher pitched strings 1 through 5 while leaving the 6th string open. This gives the relative equivalent of drop C, one whole step higher, for a version of drop D. This gives us a way to play the key of D with a strong V chord – the five capoed strings spell out an A major chord.

The Singer Calls the Key

Back in my earlier slack key days I played solo instrumentals, occasional instrumental duets, and only very rarely played with vocalists. When I started backing up singers and trying to sing myself, things got a lot more complicated. Suddenly I had to play in any key, because the vocal range is usually very particular, and different for each singer. My sudden interest in the key of D came about when I found several songs I preferred to sing in that key.

There are other reasons to experiment with tunings and cross tunings. Different tunings might lead to different arrangements, different emotional colors. Whatever the reason, whatever the style, I hope this little exploration of the outer boundaries of open G tuning was useful.

I haven’t used the soundcard (Realtek HD Audio) in my current aging desktop, I’ve always had an audio interface of some kind installed. I’ve used an M-Audio PCI card, a LynxTwo, an M-Audio Firewire box, an Echo Audio Firewire box, and now an RME UFX connected over USB. With the earlier interfaces there was no convenient way to record through the Realtek and interface at the same time, but the UFX has the ability to record direct to a USB flash drive. So this seemed like an opportunity to compare a well respected interface with the built-in soundcard of the PC.

Getting Hooked Up

Since I would be going into the line input of the Realtek a preamp would be required, so I fired up an M-Audio DMP3

and ran it into my handly little Coleman Audio LS3 switchbox.

From there I connected two stereo outputs, one to a pair of line inputs on the UFX and the other through a chain of adapters to the Realtek line input.

Software Mixing

My first attempts were dreadful. There was a huge difference between the sensitivity of the Realtek line input and the RME inputs. It took a bit of surfing to find a solution. By opening the Windows 7 System Tray

and right clicking on the volume control I got this menu

where selecting Recording devices took me to this list:

Double-clicking on the Line In device got me to the Properties dialogue box for this device:

which contains a tab labeled “Levels” that let me turn down the sensitivity:

The default setting appears to be a numeric value of 50, without any indication of what that means. A level of 0 cuts off the input, but a level as small as 3 gave a sensitivity similar to the line in on the RME UFX.

Cutting the Clips

With that out of the way I ran through my usual procedure, print a test tone, play a little guitar, adjust volumes in post, and render out the sample clips. One little wrinkle here, I bought the Studio Six AudioTools app for my iPhone and used that to generate the test tone:

So here they are, two short stereo .WAV files for you to compare:

V
W

As usual, I strongly recommend that you first confirm an audible difference with an ABX comparison. If you can identify the clips correctly 13 times out of 16, you can be fairly confident that you are hearing a difference. If so, what do you hear? And then, how significant is that difference to your recording goals?

Please understand that I’m not saying we should all stick to our soundcards for recording. After all, I don’t use my soundcard. But the reason is not because I hear huge audible improvements, it’s because of workflow and features. I often record more than two channels, for instance. And the software mixer that accompanies most interfaces adds a lot of flexibility. Driver options for the Realtek are limited, as is the quality of the connectors.

But if you’re just starting out recording on your computer, and you already have a mixer or a preamp and some mics, you can start recording with nothing more than a couple of adapters and some software. And the sooner you start recording the sooner you’ll start making good recordings.

The first thing to say loudly is that YouTube is a rapidly changing environment. In five years we’ve gone from heavily compressed SD to 1080p HD and more. The audio has improved from 94 kbps AAC audio to a recent high of 152 kbps. The future will see new upload formats, new transcoding tools, new download formats and bandwidth. So anything in this post is subject to change at any time.

Everything Gets Recompressed

As far as I know, YouTube always transcodes (recompresses) every upload. This is their stated process and I’ve never seen evidence otherwise. Folks who try to choose a format that avoids transcoding are going in the wrong direction. Instead, it makes sense to me to search for the most recent guidelines from YouTube and follow them. Here’s the YouTube help page “Video Encoding”. There’s a lot of information for specific video editors, a list of supported formats, and more. But I would boil the basic advice down to this – upload high quality video and audio.

Good Video Gets Good Audio

Since we’re focusing on audio quality here, let me start out with an important bit of knowledge that might seem contradictory. Your video format can limit your audio quality. YouTube delivers video downloads in several formats and sizes, but the largest format it will provide is the one you uploaded. So if you provide a 360P or 480P file YouTube will offer nothing larger than that for download. More significantly, the audio in that download will only be 129 kbps AAC compared to the 152 kbps delivered with higher resolution video formats.

In order to get the maximum quality audio with your YouTube download, your upload should be at least 720P (1280×720 pixels). The quality of your audio should be as high as possible, preferably PCM (WAV or uncompressed) if possible, otherwise a high bit rate MP3 or AAC. This results in large files and long uploads, but the current YouTube uploader has been quite reliable for me, and the upload can take place in the background or unattended.

To give a more concrete idea of what I’m talking about, here are the statistics on one of my clips, Uncle Sonny Chillingworth’s Slack Key #2 (Mahina’s Trot) http://www.youtube.com/watch?v=RjV1zzM6RDM posted at my slack key channel. http://www.youtube.com/franguidry

We can see three levels of audio quality along with the five levels of video. The 1080 file delivers 10 times the video data contained in the 240 stream, so we need a pretty high speed connection to view these files without buffering and pausing. The 720 stream is the smallest video format that delivers the 152 Kbps audio. For most of us the 720 stream is the best choice, since it’s more likely to stream smoothly but includes the highest quality audio.

What Do You Hear?

That brings up a fascinating question – how good is YouTube audio? It certainly has been terrible in the past, at least my results were terrible and many other people complained. Have the recent upgrades improved the quality enough to match the quality of the original file we uploaded? If you’ve read some of my other posts about comparing audio, you know what I’m going to say: ABX! Listening tests that are not well controlled are worse than no test at all, because they convey bad information. Good listening tests are level matched, same performance, and double blind.I’m going to make a couple of files available for you to compare, one extracted from the video I uploaded, the other from the same video after YouTube processed it and I downloaded it. In the past I’ve sent the comparison key in response to any post or email requesting it, but this time I’ll only send the key if you convince me that you performed an ABX test on the files.

What Is ABX?

ABX is a testing technique for confirming audible difference between two test subjects. The two subjects are A and B, one of them is presented as X, the unknown. The listener identifies X as either A or B, and repeats this evaluation enough times to eliminate the likelihood of successful guessing. For hardware devices it requires a lot of trickery and trouble, but for digital recordings ABX is a snap. There are free ABX tools for both the PC and Mac. This post at the Hydrogenaudio Forum describes ABX in more detail and explains why a number of trials is necessary. This Wiki article explains why the rule of thumb for a demonstration of difference is 13 successes out of 16 trials.

I created a blog post and a video explaining how to download and use the foobar2000 ABX tool. Please note that my current advice on the number of trials would be a minimum of 16 trials with at least 13 correct answers.

Mac users can download ABXer, another free utility that manages ABX sessions for you.

For Your Listening Pleasure
So here are a couple of files you can download and compare:

88.wav
99.wav

These are the audio tracks from a video I posted recently on my slack key channel, Uncle Sonny Chillingworth’s Slack Key #2 (Mahina’s Trot). I shot this video on a Panasonic Lumix GH2 and captured the audio using a Sony PCM-D50. I tweaked the audio a bit when I edited the video in Edius and exported a Blu-ray format video clip with a 48 Khz 16 bit stereo PCM audio stream (I used 48 Khz because that is the default in Edius, not because I think the audio quality is any different from 44.1 Khz). This is an uncompressed format, like WAV or AIFF and requires a bit rate of 1536 Kbps, quite a big pipe. When I captured the YouTube 1080 download, the audio had been compressed to a 152 Kbps 44.1 Khz stereo AAC stream. The audio is 1/10th the size after YouTube processes it. That audio compression is very useful, because it delivers smooth streaming with a slower connection and it leaves more bandwidth for the video side. But does the result deliver satisfactory audio quality? I used FFmpeg to extract the audio from the original upload clip. I loaded the 1080 download MP4 from YouTube into REAPER and rendered the audio stream as a 48 Khz 16 bit PCM stereo file, matching the one from the upload clip. Of course, REAPER also uses FFmpeg to manage video. Now it’s up to you. Download the files, they’re big so be prepared for a wait. Download and install the ABXer for your OS. Take 16 tries at telling the files apart, and let me know how it goes. Can you tell the difference between the orignial and the file with 10 to 1 compression?

The reason I need your help is that it is impossible for me to prove the files don’t sound different. I can only demonstrate that I personally can’t hear a difference, and I’m not very motivated to hear the difference. I’m biased. Someone who is confident that a difference exists is likely to try harder, listen more intently and longer, to find that difference. So if a few of you can demonstrate that you hear a difference, I’m going to agree with you.