Q: What is it?
A: Comb Stereo is a way of
sending a stereo signal through a mono channel. Sort of. (It has limitations...)
Q: Is it real stereo?
A: Yes. It does send two
channels separately, so it's real stereo. Sounds can be sent in the left
or right channels in any combination.
Q: Is it good stereo?
A: Well.. No! Don't expect it to be as good as DAB, DRM, FM Stereo, AM Stereo (C-QUAM), or even ISB stereo (perhaps the best suited for shortwave, but not currently used).
But the other systems have problems...
DAB uses digital modulation and is transmitted on VHF. This causes reduced range compared to FM, and many services have a lower than ideal bitrate causing digital artifacts which reduce quality.
DRM (Digital Radio Mondiale) also uses digital modulation and is mainly used on Mediumwave (AM Band) or HF (shortwave). DRM can drop out (go silent) due to severe fading, severe noise, or interference. The bitrate is low, and this causes digital artifacts which reduce quality. There is a tradeoff between error correction and bitrate. Making the signal more error-proof reduces the bitrate and reduces quality.
FM Stereo causes a 23dB degradation in SNR compared to mono. This causes a big increase in hiss levels. FM Stereo also uses a much wider modulation bandwidth of 53kHz, which makes it more susceptible to multipath degradation than FM Mono.
C-QUAM AM Stereo suffers from "Platform Motion" during fading conditions, which makes the sound wander left and right. The SNR is slightly lower than mono AM.
ISB Stereo suffers from something similar to C-QUAM, in that selective fading changes the stereo balance - which also makes the stereo wander left and right. The SNR is slightly lower than DSB AM, because the sidebands are not coherent (in-phase).
Comb Stereo does not have these problems!
The bandwidth is the same as mono - So the SNR should be about the same as mono.
Selective fading doesn't affect the comb bands much, so the balance is largely unaffected by selective fading notches.
The Comb Stereo artifacts
are much like typical music effects of echo, chorus, fast reverb or room
reflections.
But Comb Stereo has some of it's own problems...
The Comb Stereo encoder is quite good, in that it's very compatible with mono listening. For the most part, there is no obvious difference in sound quality.
BUT.. The basic Comb Stereo decoder has a problem of splitting the sound into Left and Right ONLY. There is nothing in common with the two channels, so there is no "centre" to the stereo image!
The enhanced Comb Stereo decoder corrects this by special Left/Right processing to reduce separation on mono sounds and boost separation slightly on Left or Right only sounds. This produces a much more normal sounding stereo decode.
Comb Stereo encoding causes
the peak levels to increase, which lowers the audio loudness. But since
it is done before modulation processing this doesn't matter too much.
Comb Stereo has some strange characteristics...
On peak reading VU meters tested using white noise, the L/R separation appears to be only about 6dB. But the actual separation is much higher.
Comb Stereo divides the Left and Right channels into a "comb" of frequencies to send the stereo information, so each channel only has half the frequencies present. But mono/centre sounds are not encoded, so all the mono/centre frequencies are sent. Also, stereo difference (L-R) is not combed either! So "surround" sounds are also flat (above 200Hz). And, strangely.. it sends the stereo difference (L-R) as if it were mono L+R! (but delayed and phase shifted) This means surround sounds decode as centre sounds, so the system can't cope with more L-R than L+R signal.
The majority of sounds in music are in the centre, so this scheme is very compatible with mono. Most sounds are not encoded, and so the encoder output sounds mostly normal. Sounds that are only in the Left or Right channels are comb filtered, and may have a slightly "hollow/pipe" sound to them. However, our ears get used to hearing the comb filter artifacts because they happen every day - anywhere you have sounds reflecting off flat surfaces like walls, ceilings and floors.
The new enhanced v7e and v7f decoders feature a special multiband stereo expander to improve performance. The result is far more natural sounding stereo with much fewer artifacts, AND good separation.
Hear a test of the new decoder here on YouTube.
Comb Stereo is amazingly simple and effective - considering it's sending two channels over a single path, with no increase in bandwidth!
* Comb Stereo was rapidly
developed from an idea in about 2 days, and is still in it's early stages.
Expect more improvements!
Q: Great! So how can I decode it?
A: Download the decoder plugin from here, and feed your receiver audio into it using your PC:
32
Bit Version 7g12 TIAMS Enhanced Comb Stereo Decoder with Auto-EQ NEWEST
TIAMS/RNEI
wide VERSION 7g12 - Updated December 23rd, 2020. For TIAMS and RNEI on
WRMI 5850kHz only.
Better stereo, more bass
separation, less artifacts, selective fade reduction, more natural sounding
bandwidth expansion, and flatter Auto-EQ.
32
Bit Version 7g0 TIAMS Enhanced Comb Stereo Decoder with Auto-EQ TIAMS/RNEI
wide VERSION 7g0 - Updated December 2nd, 2020. For TIAMS and RNEI on WRMI
5850kHz only.
32
Bit Version 7f7 TIAMS Enhanced Comb Stereo Decoder with Auto-EQ TIAMS/RNEI
wide VERSION 7f7 - Updated October 8th, 2020. For TIAMS and RNEI on WRMI
5850kHz only.
Or, use the combined AM Sync
Demodulator and Comb Stereo Decoder! (for 0Hz IF IQ input, like KiwiSDRs):
AM
Sync Demod plus Comb Stereo Decoder, DNR and Equalizer OLD VERSION
(Enable
"Comb Stereo Decoder" in settings)
* If the encoder is upgraded,
you will most likely need a new version of the decoder. Check back here
for updates!
Currently Encoder v7
is used. Any v7 Decoder will work, but v7g12
is the latest and best.
Please note:
Q: How can I run those
plugins in real time?
A: Try VSThost:
https://www.hermannseib.com/english/vsthost.htm
Here's a guide from RNEI
to setting up VSThost for Comb Stereo decoding:
https://rnei.org/stereo
Q: Who's using it? What programs
can I hear in stereo?
| https://rnei.org | RNEI on 6070kHz from Channel292 Germany AND NOW ALSO 5850kHz WRMI Florida! |
| https://twitter.com/ThisIsAMusicSho/ | This Is A Music Show on 5850kHz and 7780kHz WRMI Florida |
Q: But I want to hear something NOW!
A: Well, have a listen to some of the newest recordings from shortwave here:
http://homepages.ihug.com.au/~daz2002/tech/KBC/index.html#tiams
Some of them are in decoded
Comb Stereo!
Q: I'm using the Comb Stereo decoder on another radio program, and it sounds like it's in stereo! But it's not listed here...
A: The Comb Stereo decoder
works similar to a stereo synthesizer, so if it's fed with non-encoded
material it may sound like stereo - but it ISN'T real stereo! Only encoded
material will give you real stereo
sound.
Q: So where did this idea come from?
A: From a chat with Roseanna
from RNEI on March 30, 2020:
Comb Stereo is similar to:
"Frequency Division Multiplex System Using Comb Filters" by Yoshimutsu Hirata, Feb. 11, 1974
- but with major improvements.