DRM Digital AM audio samples

[k4kyv]

Unlike the Ibiquity all-digital system, which occupies a bandwidth of 20 kHz, this new system should have a bandwidth of only 10 kHz, and thus finally put to rest the nighttime first-adjacent interference problems that plague the band. Even with this narrow bandwidth, the DRM system can provide a bit rate in the neighborhood of 25 kbps, and offer very reasonable audio quality. Check out some of the audio samples on the DRM Web site.  The page allows you to make direct A/B comparisons between analogue DSB AM and DRM at the click of a button.



http://www.drm.org/system/audiosamples2.php

[Bacon, WA3WDR]

Sounds pretty good, especially at 22-25KBPS.  10K RF bandwidth?  Not bad.

[W1DAN]

Don:

I built a KE1GF converter, works great.

BTW, DRM occupies the full channel as compared to IBOC which allows a narrowed analog AM signal to co-exist. So you will not see it on terrestrial AM for the "transition period" at least.

Met the CE of Radio Canada who is a big proponent of DRM. They transmit DRM from Sackville every week. DW aires DRM at around 4mc.

I do not think DRM will fully be adopted...receivers are much more expensive than analog ones.

Also check out this month's QST about an article on HamDRM...digital audio in a SSB bandwidth.

73
Dan
W1DAN

[k4kyv]

I read the QST article.  As would be expected, they are advocating the magical bandwidth equivalent to SSB, but I am not sure that the audio would be limited to SSB quality. I wonder what bandwidth would be required for audio response out to 5000~.  The Deutsche-Welle signal on 3990-4000 is exactly 10 kHz wide, but I don't know what audio response is specified for that signal.  Since the signal density of the DRM signal is much greater than that of analogue AM, I suspect it would cause much greater interference than would a normal AM signal of the same signal strength and bandwidth.  I seem to recall that a certain s/n ratio is necessary for the full DRM quality, so while it may work for strapping signals, it would be of limited use for weak-signal reception. I suspect DRM reception is similar to the capture effect of FM, with the audio abruptly dropping out below a certain signal to noise level.

Supposedly in the AM broadcast band, a 1 kw conventional AM signal equivalent would be 250 watts DRM.  I am not sure if this is in terms of interference potential, or  effectiveness of the signal.

The A/B comparisons on the DRM web page are quite impressive, especially the Portugal to Germany link.  With the analogue signal you can hear the rolling fading effect so typical of shortwave AM, but DRM seemed to be FM quality (as best as could be determined by listening via MP3 audio).

Until receivers become much less expensive, I doubt that DRM will get off the ground for broadcasting services in third-world countries, which, except for the US-based Holy Rollers, is about all that is left on the Shortwave broadcast bands.  But broadcasting interests are likely to play the DRM card in an attempt to hold on their HF allocations, particularly on 40m, at the next world radio conference.

[Rob K2CU]

I have been experimenting with WinDRM for about two weeks now. I have yet to receive anyone well enough to get the data to decode. I have done the computer to computer test to verify that all the software is working. The audio works well and sounds as good as the microphonium you are using. There is a noticeable delay, about a second or more. I am looking for anyone else in the Northeast interested in trying it. It works with the same computer sound card to radio interface that is used for other digital modes like PSK-31.  It would be interesting to run some frequency response and audio dynamic range tests, besides just talking. One thing that is neat is the visual ID that is done by creating an audio burst that makes your callsign appear in the waterfall display of the listening station.  And it does fit in a 2.5 KHz bandwidth.

[AB1GX]

This DRM stuff sounds like a lot of fun.  Certainly more fun than blowing up tubes on my AM rig.

Where do I look up the technology?  Is is binary PSK?  So can I just down convert the signal to near baseband, decimate with an A/D and feed the resulting digital signal into a processor to do DSP (matched filter, Interpolator, phase recovery)?  Then there's probably some decompression like mp3.

Now the big question... How does one phase modulate a Class E/F rig???

[Rob K2CU]

See:

http://n1su.com/windrm/

You need a fairly clean SSB rig for this and other digital modes that use the PC soundcard.

[W1DAN]

Don:

The DRM or HamDRM audio bandwidth is not nexessarily limited to the occupied RF bandwidth.

In bit rate reduced audio, many tricks can be done to reduce the resulting data rate and thus occupied RF bandwidth. Usually perceptual coding is used that reduces the bit depth and sample rate, as well as eliminates softer audio near louder audio's frequencies. So it is possible that DRM and Hamdrm can pass 20khz audio. Usually, the sample rate is limited to 22.050khz or lower to provide quick and easy data loss.

BTW DRM is capable of passing stereo (I have heard DW in stereo on 4mc),

IF phase shift as well as interference easily kills these streams.

73,
Dan
W1DAN

[AB1GX]

After looking at the spec, the am signal has both the analog and a digital subcarrier.  With so little spectrum for the digital, there's good reason that the signal is so sensitive to fading.

I'd like to see the analog dropped and have digital fully occupy the 10kHz bandwidth.  That will help a bunch with fading.  It can be called narrowband spread spectrum!  Everyone uses the same feedback shift-register.  Also, since all available channel bandwidth is used, the SNR will be much better than with normal AM (or SSB - which I'd like to see become vintage).

Tom, AX1GX