Improving sound quality without changing hardware. DIY.

Status
Not open for further replies.

robgmn

Member (SA)
I got a lot of much-appreciate help from some folks in a thread I created about getting a box working. I hope sharing this info might "pay it forward" a little for those who really like to tinker with audio,

Finally got a 'box I've wanted since I was a teen (Panasonic RX5150).
My ears are a little more discerning with age, and I was thinking about ways to make it sound better.
I wanted to keep it stock, so my only option was to digitally EQ the music it was playing.

This will not work for FM (unless you have a whole-property transmitter, which I do) but can be done with mix tapes, digital music (via a line-in), or even with homemade mix CDs. It seems cumbersome, but I'm a tinkerer and it really only took about an hour of time (I've burned hours on things a lot less fun). 'Something to do on a rainy day.
Plus, I'm not a collector, so I only need to do this once.

What you'll need: dB meter (or smartphone with calibrated mic and free dB meter app); Audacity (free download); JRiver Media software (free trail) or similar software with a parametric EQ option (31 band EQ option is next best); 1/8"-to-RCA audio cable (or a blank CD to burn your own test disc)


Process
1. Use Audacity to create your own test tracks. Click "Generate>Tone" to create a sine wave ten seconds long of each of the following frequencies with a 5 second break between each tone (create silence by clicking "Generate>Silence"): 50Hz, 63, 100, 125, 250, 500, 700, 1K, 2K, 4K, 6K, 8K, 10K, 12.5K, 15.5K, 17.5K.
Export to a wav file if you want to burn a test CD.


2. Play the test tones through the 'box (using the audio cable or your burned CD) and measure each frequency with your meter. Create your baseline volume setting by starting with 10KHz. Set the box volume so the 10K tone registers at least 3dB louder than background noise on your meter.
Write down the results.

3. Go into your music software and open the parametric EQ. Pick the "quietest" frequency in the 60-250Hz range from your written results. This is the baseline you'll use to adjust the others ('tis better to reduce than to boost when it comes to EQ).
Using a Q of 2, set negative gains for the frequencies that go above the "target" by more than 1 dB.
If two frequencies right next to each other need to be reduced, split the difference ( ii.e. if 250Hz and 500Hz both need a 2dB reduction, do a 1.5 dB reduction each).
Remember to allow for a normal high-end rolloff.

4. Apply the EQ to some music you are very familiar with (as played through a quality audio system) and play it through the 'box.
You can also apply it to the test tones and measure the difference it makes.

If you are satisfied with the changes, you can apply it to CDs that you burn for 'box use, apply it to tapes as you record them, or to all music you put on a digital player (I have a dedicated old phone I can use as a source for the Panasonic).

i'd bet a nickel there is a tinkerer out there who would put a miniDSP inside a box and apply it to the signal chain.

What surprised me about this test was that (despite what I've heard about boombox tweeters), my spectrum analyzer was picking up output from the Panasonic all the way up to 20Khz. It was low level, but it was definitely there.
What didn't surprise me (based on what this this sounds like) was the sloppy upper mid-bass. It dropped 5 dB between 63 and 125Hz, shot up 8dB between 125 and 250Hz, then dropped 6dB between 500 and 700Hz. Another problem area was a 7dB dropoff from 4 to 6KHz. From 6KHz on up, the roloff was actually pretty consistent.
Smoothing that mess out made a huge difference.

I'll try to post a picture of the spectrum analysis plot if I can.
 

Attachments

  • PANASONIC SPECTRUM ANALYSIS.PNG
    PANASONIC SPECTRUM ANALYSIS.PNG
    19.5 KB · Views: 3

Reli

Boomus Fidelis
Cool stuff

If I'm playing stuff thru a phone, I'll make a V-shaped EQ curve on the phone's settings menu, because most boomboxes have way too much midrange and mid-bass. Especially the cheaper boxes. You can see it in the service manuals, the best box I've seen had a frequency response of 30 to 18,000 Hz (when using Metal tape), while the cheapest boxes had frequency responses of 100 to 8,000. Whether that data is accurate or not is besides the point, the point is that the cheaper boxes need a lot of help.
 

robgmn

Member (SA)
Reli said:
Cool stuff

If I'm playing stuff thru a phone, I'll make a V-shaped EQ curve on the phone's settings menu, because most boomboxes have way too much midrange and mid-bass. Especially the cheaper boxes. You can see it in the service manuals, the best box I've seen had a frequency response of 30 to 18,000 Hz (when using Metal tape), while the cheapest boxes had frequency responses of 100 to 8,000. Whether that data is accurate or not is besides the point, the point is that the cheaper boxes need a lot of help.
This one wasn't "cheap" when new (I think it was in the $300 range in 1980-something), but it definitely needs some help!
 

Reli

Boomus Fidelis
Those are cool looking boxes! A member here painted his with a fresh coat of platinum, and painted the grills dark silver.

If you put your ears up to the tweeters can you hear anything? I owned 2 of those models and I couldn't hear anything from the tweeters. Maybe because I can only hear up to 14-15K.
 

Jorge

Member (SA)
Funny thing: when I listen to Bee Gees or Yello thru boombox I LOVE what I hear: it sounds as it sounded when I were my younger self when I listened to this crap. Through my hi-end it sounds totally unrecognizable... But what Reli said explains why mid is wired out of phase in RC-550: to make this "dip"!!!
 

robgmn

Member (SA)
Reli said:
Those are cool looking boxes! A member here painted his with a fresh coat of platinum, and painted the grills dark silver.

If you put your ears up to the tweeters can you hear anything? I owned 2 of those models and I couldn't hear anything from the tweeters. Maybe because I can only hear up to 14-15K.
I got lucky in acquiring it. For some odd reason, the bidding didn't go anywhere near what it normally would for this particular unit.
Superduper (who painted his box) is primarily the one who kept pushing me in the right direction during my repair quest.

The tweeters are definitely working. I've put my ear to them AND covered them during playback to confirm, plus there's no way that the "woofer" could output anything in the 18KHz range that would be picked up by my meter.
The signal was definitely there despite their 200 Ohm impedance!
 

trippy1313

Member (SA)
That "V" or sometimes "U" must be pretty common for equalizer settings huh? I do the same thing on most of mine.
 

Lasonic TRC-920

Moderator
trippy1313 said:
That "V" or sometimes "U" must be pretty common for equalizer settings huh? I do the same thing on most of mine.
I see many EQ settings on people boxes, some in a V, some in a U and some just SMASHED all to the top which to me just boggles the mind that that could sound good at all.

This is a very interesting thread. I have never tried anything like this, I did think about trying to use software to re-EQ tracks, but this is way more scientific.
 

robgmn

Member (SA)
I'm playing around with the "next step" which is to have a system-based parametric EQ app added to the phone that I use as my music source.
That way I can listen to music from memory, Spotify, Pandora, or stream from my home server (via Muzecast), and have it all EQd roughly the same.

I have a habit of getting an idea in my head and hanging on to it like a pit bull :-)
 

hopey

Member (SA)
Hi robgmn, how did you measure the output frequency from the box? The low end featured cannot be produced by the radio maybe you are measuring the MIC pickup including noise. You must exclude the known unknowns.

Trust your ears by adjusting the eq on the playing device.
 

robgmn

Member (SA)
hopey said:
Hi robgmn, how did you measure the output frequency from the box? The low end featured cannot be produced by the radio maybe you are measuring the MIC pickup including noise. You must exclude the known unknowns.

Trust your ears by adjusting the eq on the playing device.
I used a calibrated mic and measured the power of each test tone played with the volume control in a fixed position, and also used an FFT spec analyzer to see that sound was being produced even when not loud enough to show on the dB meter (I didn't want to crank it high enough for 18KHz to register for fear I might cause damage when playing lower frequency tones).

Background level in the "test environment" was about 49dB.
The low end shown CAN be produced by the box, just not at a volume that would make it useful without severe distortion (likely amp clipping before cone breakup occurs).

A small speaker can vibrate slowly with no problem, it's just that it can't move much air in doing so which makes it essentially useless as a low freq. transducer (though if it's in your ear canal, even a 2mm driver can produce very audible 20Hz tones).
Notice that even the 50Hz tone is a full 18dB below the 500Hz tone on the 'box. Massive difference there.

I trust my ears to know that without EQ the sound is pretty bad at best. I also know that the ear is easily deceived when trying to EQ properly "by ear".
For fun, try this experiment: take a 1/3 octave EQ and alternately fully boost or fully reduce sliders from left to right. Switch it in and out of the chain and compare the sound (be sure to adjust the gain of the EQ'd sound to match non-EQ'd). Interesting?

Since the only EQ on the box comprises bass/treble/loudness, this box sounds like crud without some REAL EQ work...
 

hopey

Member (SA)
It would be interesting to perform the same profile on another boombox for comparison. I think however your expectations are too high, after all these are portable and have high spec but limited range speakers which suit the battery life. You could replace the Drivers and extend your range but reducing battery run time.
 

robgmn

Member (SA)
hopey said:
It would be interesting to perform the same profile on another boombox for comparison. I think however your expectations are too high, after all these are portable and have high spec but limited range speakers which suit the battery life. You could replace the Drivers and extend your range but reducing battery run time.
Replacing drivers is not really an option (as mentioned in my o/p - "I wanted to keep it stock, so my only option was to digitally EQ the music it was playing").
If there were other enthusiasts in my area, I'd be more than happy to measure their rigs to see how they perform.

When weather gets cold again and I prefer to be indoors, I'm going to do some experimenting with standalone mini speakers I have modified to see what can be done there.
In some instances it takes very little change in EQ to make a very positive change to the sound. The small changes listed below when applied to a specific desktop speaker make it from sounding "good" to making it sound far outside of its price class.
The EQ changes that I have listed in my o/p alter the character enough that it sounds like a different boombox altogether.

Filter 1: ON PK Fc 67.7 Hz Gain 1.3 dB Q 2.0
Filter 1: ON PK Fc 140 Hz Gain -2.0 dB Q 1.0
Filter 1: ON PK Fc 345 Hz Gain -2.9 dB Q 0.3
Filter 1: ON PK Fc 675 Hz Gain 2.8 dB Q 3.0
Filter 1: ON PK Fc 870 Hz Gain -1.9 dB Q 3.0
Filter 1: ON PK Fc 1235 Hz Gain -5.1 dB Q 2.0
Filter 1: ON PK Fc 1785 Hz Gain -1.1 dB Q 6.0
Filter 1: ON PK Fc 2235 Hz Gain 1.3 dB Q 4.0
Filter 1: ON PK Fc 3080 Hz Gain -1.2 dB Q 4.0
Filter 1: ON PK Fc 4500 Hz Gain 2.0 dB Q 2.0
Filter 1: ON PK Fc 6300 Hz Gain -3.4 dB Q 2.0
Filter 1: ON PK Fc 7000 Hz Gain -2.8 dB Q 2.5
Filter 1: ON PK Fc 9000 Hz Gain -2.2 dB Q 0.3
Filter 1: ON PK Fc 9700 Hz Gain -1.7 dB Q 3.0
Filter 1: ON PK Fc 14000 Hz Gain -2.5 dB Q 0.8
 
Status
Not open for further replies.