Sunday, July 23, 2017

Visualization of Audio Loss in Earbuds

When you listen to music from your phone, tablet, or mp3 player, do you ever notice that you keep pushing the earbuds in to get that bass that always seems to be missing? Have you ever switched to headphones (that sit over the ear) and notice how much fuller the music sounds?

Earbuds versus Headphones


Before we get started, we should probably make clear the difference between earbuds and headphones. Earbuds refer to the hearing devices that are inserted into the ear, and thus have a very small speaker. Headphones refer to the padded, speakers that sit outside and slightly away from the ear. (Walkman-style headphones are somewhere in the middle, but not really considered here, as we're looking at the extremes.)

Frequency Range versus Apparent Frequency Range, and the Human Ear


Frequency range refers to what frequencies (low to high) an audio output device can emit. Apparent frequency range, however is how your ear hears those frequencies. For example, is the bass too quiet? Are the highest frequencies a little louder? This is really the volume at which a frequency (or more likely a range of frequencies) is perceived.

The human ear is, in youth, capable of perceiving sound from roughly 20 Hz (very low bass sounds) to 20,000 Hz (very high-pitched sounds). Starting in our mid teens, we begin to gradually lose the ability to hear the higher frequencies, as the hairs in our cochleas begin to die. (Interesting side note, we can perceive sound lower than 20 Hz, but we don't hear it, rather, we feel it as vibrations.)

The Problem with Earbuds


Earbuds, as with most audio-emitting devices, are capable of the full range of human-perceptible sounds from 20 - 20,000 Hz. The problem is, those sounds drop off (get much quieter) on the very low, and very high ends. The apparent frequency range becomes smaller.

Conducting a completely non-scientific study, I went out to the Internet looking for specs of various models of earbuds, and headphones. I looked for areas where the volume dropped off on the low end, and on the high end of each, and averaged the results from the earbud category, and those from the headphone category. Here's what I found:

Frequency Ranges
Headphones (average of 5): 28.75 Hz - 14 KHz
Earbuds (average of 5): 176 Hz - 8.2 KHz

From this we see the range is narrower, with earbuds producing less volume of frequencies below 176 Hz, and above 8.2 KHz.

Then, I though about how I might visually demonstrate this loss of fidelity. As it happens, the light we see occurs on a band of perceptible frequencies, as well. With just a little effort, we can map visible light against a spectrum of perceptible audio:

Mapping visible light to audible sound. Note that the measurements increase in different directions. This is due to how each is typically measured. Light is measured in the width of a wave (in nanometers), and sound is measured in the number of waves per second (Hz). In both graphs, the longer waves are on the left, as are the fewer number of waves per second.


Next, we can identify which colors would be missing from an image that is otherwise, fully visible. Using the frequencies above listed for earbuds, we see that much of the red will be missing, as will all of the violet, and darker blues. We need to locate images with a broad spectrum of color, and deep color saturation to see the real effect. I found two that seemed appropriate:


In the image above, the top photograph (analogous to the headphones) shows the full spectrum of colors in the flowers. The bottom picture (analogous to earbuds) removes most of the red (low end of the spectrum), and all of the blue & violet (high end of the spectrum).


The next image shows the same thing, but is a bit more pronounced, as the blue feathers of the parrot are a significant part of the image. The left image is like the headphones, and the right, like the earbuds. Again losing the high frequencies (blue/violet), and lowest frequencies (reds) leaves a washed out image, much like earbuds provide a washed out version of the audio.

Is it a Problem?


Now interestingly, under certain circumstances, this is completely irrelevant. For example, if you are listening to your favorite podcast, there really isn't a need for the higher and lower frequencies, as the human voice when speaking occurs around 300 Hz. Using our visualization process above, it would be the equivalent of starting out with a photograph of grass. Since it's all green anyway, you're just not missing that much. Switch to music, however, and you end up missing much of what the artist intended for you to hear.

Saturday, July 1, 2017

Backing Tracks for Guitar

Little by little, I'm finding spots of time here and there to start playing again. The left-hand fingertips are getting rougher, and with YouTube, I have a great source of lessons on how to play nearly everything.

I don't know why I had not thought of it before, but I decided to do a quick search on backing tracks for guitar. There are lots of sites out there with thousands of tracks. I've made a list here (as I'll probably forget them in the morning). I've also provided a few notes on each, in case you're looking for something specific.


Some of these (GuitarBackingTrack.com, GuitarVoice, and the Backing Tracks Wiki at Reddit are community-sponsored. You can improve the collection by uploading your own tracks.


Update: 14 Aug 2022: Removed dead links.

Sunday, December 25, 2016

Oscilloscope Project

I've found a use case for an oscilloscope. Dad has a nice, industrial scope from the 1960s, but being the size of a weekender suitcase, and weighing just as much, It was a little more than I care to have lying around the house. The next solution is the oscilloscope app for Android (of which there are quite a few). But when I found this one, I just couldn't resist. I picked it up from China (direct), and it only cost $21.00 plus shipping. Hard to beat that price, and to make things better, it's a kit, so I get to build it. This scope is a JYETech DS0138 scope.

Tools


I ended up using the following tools:

  • Soldering iron (instructions call for a 20w. I used a 30w, and it worked just fine - I just took care not to linger, especially when working around the processors.
  • Solder - good to have two sizes. I used one that was ~1mm in diameter, and the other ~2mm. The larger works really well for most components. The 1mm size was really useful on the USB and slider switch leads as they are really small, and close together.
  • Multi-meter - I use a Fluke, but for this project, just about anything would work.
  • Small Phillips screwdriver
  • Wire cutters
  • Small needle-nose pliers
  • Solder remover (optional) - This should really be part of any project to aid in desoldering, if you make a mistake)

You will also need a 9v power supply (center positive), and for testing, something that can generate a signal (a cell phone with a tone generator app works really nicely), and a male-male 1/8" audio cable.


Building


Step one is taking stock of the contents (building kits is time consuming enough - no need to get knee deep into the build phase, only to find out there's a piece missing). Everything was there. There are two kits like this available from various Chinese resellers. The first is strictly the oscilloscope (board and LCD display). The second includes the Plexiglas case for $4 - $6 more (I opted to get the case, also).



One of the things to watch out for when inventorying components is that the resistors are VERY small, and they are all 5-band coded. This means it's very difficult to read the colors to validate the correct values. I ended up measuring them all with my multi-meter, rather than trying to read the color codes.

Next, we start soldering components. As a rule, secure resistors, diodes, jumpers and small capacitors first. These tend to sit lower on the board, and are more difficult to reach if there are larger components surrounding them. Next, hit the larger components on the inside of the board, and finally larger components and connectors on the edges of the board.

Once complete, you'll have a nearly finished board like this:



At this stage comes testing. This is really important. The instructions have you perform voltage tests before soldering over a set of jumpers. Important to follow the instructions so you don't damage the processors or display in the event something is not wired correctly.

Finishing


This was actually the hard part. You're provided 4 bolts with 8 standard hex bolts, and 4 rounded nuts (for the top). Use the 8 hex nuts to set the heights of the boards, and the 4 rounded nuts to secure everything together.

The case was shipped with three pieces of Plexiglas that were smaller than the two top/bottom pieces. You will need all three pieces. The one with the most holes goes under the display, and the display bolts on to it. The other two act as spacers (filling up the empty space between the display board and the top cover).

The time comes in making fine adjustments to the heights of each hex nut. If they are too low, you risk cracking the Plexiglas. If too high, the buttons will not stick out of the top. While yo are working on this part, take care, as the Plexiglas is rather brittle, and can crack easily.



Once you get the height set, you can start adding the Plexiglas layers. After the three inside layers are in place, drop in the red switch extensions (they must come before the top layer). Bolt on the top, and you're finished.



I added some labels to the finished product to make it easier to read.

Sunday, August 14, 2016

1967 Fender Deluxe Reverb Restoration Project - Part 3

I've completed a few more items on the restoration. Before I jump in, if you are hitting this page first, you might want to check out PART 1 or PART 2.

Power Rewiring


OK, So one of the goals is to get the amp back to as close to the original as possible. There is a very real problem with this goal that I just could not ignore any longer. The original wiring called for a two-pronged plug with a polarity switch on the back. This is unfortunately unsafe, in that with the wrong polarity, you can easily shock yourself pretty good. So, in the interest of safety, I'm restating the goal to get the sound as close as possible to the original, but with safety first. As luck would have it, this rewiring job has no impact on the quality of sound, so here is the new power wiring arrangement:



I'm no electrical technician, so I'm not even going to begin to try to walk you through what I did (better to get the knowledge from someone that really understands this, and does it for a living). In short, however (no pun intended), I added the three-pronged cable, disabled the polarity switch, removed the Death Cap (yeah, you'll want to read up on that one), and brought the fuse inline with the hot lead. If you want to know step-by-step how to do that, check out Greg Perrine's website, Recycled Sound. He does a great job providing videos, reference documentation, and parts for your repairs.

Dating the Amp Redux


So as I was watching one of Greg's videos, he pointed out an interesting piece of information that I was not aware of. In the image above, there are some numbers stamped on the chassis (I circled in red to help you locate them easily). This is [drumroll] the manufacture date! Woohoo! So, from this number we look at the last 4 digits. They tell us the amp was assembled in the 22nd week of 1967. I was but a wee lad at only 6 months old when this amp was originally built. One more mystery cleared up.

New Speaker


I happened to locate a 12" Fender driver in Fort Wayne last weekend, and dropped a few bucks for it. It is a real fender, and it's in great condition. That said, it's neither the original, nor does it match the original Jensen specs. Quite honestly, I rather prefer the growl of the Peavy that used to be in there, but I have both, and since there's an External Speaker jack on the chassis, who knows? Maybe there is a second cabinet in its future.


Monday, July 11, 2016

1967 Fender Deluxe Reverb Restoration Project - Part 2

This is a continuation of a restoration project I undertook for a 1967 Fender Deluxe Reverb amp. You can find Part 1 HERE, and Part 3 Here.

New Tubes!

So after a bit of a break, I got back to work. My dad (a retired electronics technician for what was once Bethlehem Steel) called me from a hamfest and said there's a guy with tubes. He asked if I needed any, and after a bit of Q and A, he was bringing home a spare 12AX7, and a matched set of GE 6V6 tubes.

 

Now, these are the originals, in the original boxes (they really are pristine), which makes them 'new old stock' (NOS) tubes. I've done a good bit of reading about these elusive NOS tubes, and although they fetch a pretty high price, it's not really clear that they improve the sound of the amp. Either way I do get the benefit of having a matched set (something the amp was lacking before).

A New Reverb Tank!

This one was a long time coming. I ended up picking up an Accutronics 4AB3C1B reverb unit. It's a two-pring unit with 8 ohms input, and 2250 ohms output. This is as close to the original that would have come with the amp back in the day. I also picked up a cable to go with it. Here are some build notes:


I installed some tie-downs along the left side to keep the wire out of the way. I can still throw some things in the back, without too much concern of tugging on the connector cable.


I ended up having to scour the Internet for how to connect these, as installing was more than trivial, and I didn't want to have to pull it back out to look as the back side of the cover to see which color was input, and which was output.




I did cut a piece of cardboard to fit under the tank. After a good bit of trolling the forums, it seems the originals produced by Fender back in the day had cardboard underneath, so this was just trying to keep in the tradition of 'as close to original as possible'. That said, since the screws are tightly affixed to both the cabinet, and tank cover, I doubt that there is much difference in the sound of the reverb with or without. (I'm working on a post about sound quality that should be out in a few weeks.)

Friday, June 10, 2016

TubeScreamer+ Project - Part 5: Seeing Sound

INTRODUCTION

A few weeks ago, I became curious what sound passing through the TubeScreamer+ would look like. Fortunately, I have two cell phones, and the cables in hand already. I used a Samsung Galaxy s4 mini and an app called ToneGen for tone generation. As the scope, I used a Google Nexus 5x with an app called Oscilloscope. I also used a screenshot app to capture the images of the scope. Between the two phones, I have the pedal connected to each phone using two cords and two adapters (1/8" stereo-to-1/4" mono). Next, I fired up the apps, and started sending a nice clean tone (about 1000hz) from the tone generator to the oscilloscope.
NOTE
For those that did missed my earlier posts during the building project, refer to a photo of the pedal below. The H/L switch under the GAIN indicates High and Low. The original Tube Screamer TS-808 used the H position. The L position activates a capacitor mod that ends up giving the output much more low-end (for a heavier rock style). The lower switch describes two different distortion methods. The left side (with the diode symbol) is the original asymmetric diode clipping style used in TS-808. Flipping the switch to the right changes the clipping technique to one that uses both transistors and diodes.

A CLEAN SIGNAL




Above is the clean signal. The peaks and valleys are symmetrical, and have a nice logarithmic bow. In addition, the distance is exactly the same from one crest to the next. These are all traits we would expect to see from a signal generated by a tone generator.

THE ORIGINAL SOUND




In the image above, we see the effect of running the same clean signal through the pedal. Notice the nice even curves now have a slight shark-fin look. The audio for this would sound significantly distorted, and somewhat high-pitched.

ADD A CAPACITOR MOD




The capacitor mod (flipping the top switch to the right) shown above shows an increased amplitude but the wave is shaped slightly differently, as well. In audio terms this would sound similar to the previous, but with more low frequencies (for that heavy rock sound).


A CAPACITOR MOD WITH LESS TONE




In this image, you can see the effect on the wave of changing only the tone control in the upper left corner. I dialed the tone down to about 50%. The wave now has a slightly more subtle curve in the crests and valleys. Audio-wise, this translates into the same low-end distortion from above but as though someone threw a blanket over the speaker. Much of the higher frequencies have been filtered out.


This was a really fun exercise. I want to do more playing to see how different clipping styles compare, and how changes in the GAIN affect the shape.

PART 1 ~ PART 2 ~ PART 3 ~ PART 4 ~ PART 5

Friday, May 13, 2016

TubeScreamer+ Project - Part 4

THE UNVEILING


All the parts came in, the last bits of solder have hardened, and the paint is dry. And here is the final product:




The design was laid down with nothing more than a pair of gold and black permanent markers. Once the design was finished, I threw two coats of automotive clear coat down to protect the ink.

It plays like an absolute dream. The extra range in tones I get from the two mods was totally worth the extra effort (both switches flipped to the left provide the closest to the original TS-808 sound).

PART 1 ~ PART 2 ~ PART 3 ~ PART 4 ~ PART 5