DCT2000/5100 Sound Levels and Compression

In reading this thread, I notice that a number of people have raised questions about audio levels at the various outputs and the proper use of the "compression" settings in the setup menus. I believe that these issues are identical for the DCT-2000 and DCT-5100 set-top boxes. Furthermore, these issue relate to Motorola's implementation of the Dolby Digital audio standards.

I've spent considerable time reading Dolby Labs tech papers on Dolby Digital encoding and decoding and I might be able to shed some light on these issues. It is important to remember that the DCT-xxxx digital cable boxes are licensed Dolby Digital decoders and, therefore, everything they do in terms of audio can be traced to the Dolby Digital feature-set. IMO, the Dolby Digital standard is remarkable in both its capabilities and flexibility. However, those things make it fairly complicated and I think Motorola may have missed one key detail in their implementation that results in many of the volume-level mismatches that show up.

The single most important thing to understand about Dolby Digital is that is capable of unprecedented dynamic range. Looking at a high-end Dolby Digital playback system with no compression, with the volume control set to official Dolby Reference Levels: Average dialog levels (from a single channel) will be reproduced at -31dBFS (31 db below maximum full scale digital signal) and the loudest peaks will be reproduced 31dB higher at 0dbFS! The softest sounds could be at the level of audibity, perhaps - 70dBFS. This recording is intended to be played at very high playback levels. Official Dolby Reference playback levels put the average dialog levels at 74 dB and maximum peaks (again from single channel) at a whopping 105 dB). This puts the average dialog at a realistic level for spoken words (as loud as if the actors were in your living room), but the loudest possible peaks are staggeringly loud. Hats off to Dolby for providing this dynamic range. This, BTW, is the "NONE" compression setting on the DCT-xxxx settop boxes and the setting used on Dolby Digital receivers with the no compression setting.

However, on lesser systems, this available dynamic range creates problems. If the loudest peaks are too loud for your speakers (or your ears), you can't just "turn it down" or the dialog levels start dropping to the point where you can't hear the actors. Dolby has recognized the need for Dolby Digital to be flexible enough for all kinds of systems: from good to bad. They have incorporated a very, very sophisticated compression system to reduce this dynamic range.

One part of the this system would be characterized as "mild" compression and results in the kind of dynamic range you might expect from a VHS Hi-FI recording. This compression mode leaves the average dialog level unchanged at -31 dbFS, but reduces the peak levels by as much as 10 dB and boosts very quiet sounds by as much as 15 dB. When played back at Dolby reference levels, the movie will still sound "loud" by those enormous cannon shots and explosions won't shred your "home theater in a box" quality speakers. This compression setting is commonly known as "midnight mode" on Dolby Digital receivers and I believe this corresponds to the "LIGHT" compression setting on the Motorola DCT-xxxx boxes.

This is STILL way too much dynamic range for signals that are being sent by RF modulators to the speakers built into a TV set. These signals will totally overload the RF modulators (for sending a signal from the set-top box to the Channel 3 input on a TV) and will drive TV speakers into gross distortion if turned up loud enough to hear the actors. So, Dolby has designed yet another level of compression into the Dolby Digital standard so that Dolby Digital can be used with these types of low-end systems. In their white papers, Dolby refers to this compression mode as "RF MODE" and it is REQUIRED in set-top box Dolby Digital decoders. This mode limits the loudest peaks to just 10 dB above average dialog levels and boosts quiet sounds to within 25 dB of average dialog levels, for a max dynamic range of just 35 dB -- about what we are used to from over-the-air TV broadcasts. Then, the boost the average dialog level UP by 11 dB to -20dbFS, so that you can hear the actors when the playback volume levels are reduced to what an average TV can reproduce. This highly compressed RF mode is not found on Dolby Digital receivers, but it is the "HEAVY" compression mode on the DCT-xxxx set top boxes.

Because the average signal level is boosted by 11 db in RF mode, you would get a huge volume difference when switching between RF heavy compression and the other compression modes on a set-top box. To prevent this, the Dolby Digital decoder is SUPPOSED to kick in an analog pre-amp level attenuator of 11dB (after the Dolby Digital decoder) when RF mode compression is activated -- therefore the average signal level will be unchanged when you switch compression modes. I'm fairly certain that Motorola overlooked this requirement and failed to implement the offsetting 11db analog attenuation stage with their RF mode. That's why there is such a HUGE volume difference when you go from "HEAVY" compression mode to either "LIGHT" or "NONE".

All of this just applies to the decoded analog signals at the RCA line outputs or the RF coax outputs. I don't believe that it applies to the optical or coax digital audio outputs on the DCT-2000 or DCT-5000. Those should be supplying a pure pass-through of the incoming Dolby Digital bitstream which will be decoded by your Dolby Digital receiver at one of its digital inputs. As near as I can tell on my DCT-2000, the digital audio outputs are not effected by the compression (or volume) settings on the set-top box.

So why are analog stations so much louder than the digital stations, even at the digital audio outputs? Well, digital stations have a Dolby Digital encoded audio track (whether it is 5.1 or stereo or mono). Dolby Digital encodes these at a fixed (and known) level with average signal levels at -31dBFS. But, the analog stations cannot have a Dolby Digital audio track -- their audio is plain old analog MTS stereo. The DCT-xxxx must run this analog stereo signal through an analog-to-digital converter and send it out the digital audio outputs as a Dolby Digital 2.0 PCM bitstream. The levels of this converted signal are clearly much higher than the officially prescribed levels of the Dolby Digital audio tracks on digital stations.

So what does this mean? If you have a Dolby Digital receiver, you should connect the digital audio output (optical or coax) from the DCT-xxxx to a digital input on your receiver. Then, make sure that you activate the digital audio input rather than the analog input on your receiver. If you want less than full dynamic range, use your receiver's compression settings.

There is no need in the above situation to connect the analog audio RCA outputs from the DCT-xxxx at all. But, if you do so, then you need to set the audio parameters on the DCT-xxxx properly. For good audio performance from a home theater surround system, you should choose either "NONE" or "LIGHT" compression settings. You should also chose the "Matrix Stereo" setting so that your receiver can decode the stereo signal properly with Dolby Pro Logic decoding.

That's the way I see it. Hope this helps!

BACK TO DCT5100 FAQ