Index to Q&A Home Page

 

Q&A # 53 - March 3, 1998

Staff

Divider

Q What is the real difference between a 1 bit and an 18 bit processor? Is it a tech term that one can really hear the differences or is it a marketing scheme? Also I noticed that some DVD players have 9 bit and others have 10 bit. What do they really mean?

A 1 bit processors usually refer to the digital to analog conversion (DAC) method used, and 18 bit, 16 bit, 20 bit, or 24 bit processing can apply either to the same thing (DAC), or signal processing (DSP) resolution. For a quick run down on digital, look at John Busenitz's recent article in Secrets. Converting back to analog, both the single bit processor, and any variations of the multi-bit, accomplish essentially the same thing, but use different techniques. The single-bit processor uses digital signal processing to evaluate the equivalent values for the multi-bit. They then provide the recorded quantized level by turning on full blast for a ratio of that sample duration equivalent to the ratio of the recorded quantized level to full output. After going through the filter, the signal loses the drastic switching part, which is a very high frequency component of the entire spectrum, and smooths out the length of samples to a relatively slower changing, and hence smoother, averaged output. In multi-bit converters, each bit is assigned a voltage level in the converter, and for each sample, the output is the sum of the voltage levels of the "on" bits.

Think of it this way. Most of our audio material is 16/44.1, meaning that as it was sampled  44,100 times in a second (44.1 kHz), each sample recorded at a signal voltage to the nearest quantized level, with 216 total levels. Each bit can have a value of either 1 for "on" or 0 for "off", i.e., two values, which is the basis of binary mathematics. With CDs, the 16-bit "word" which represents the code for a particular voltage can be any combination of sixteen 1s and 0s. There are a total of 216 combinations of words that are 16 bits in length, and each bit is either a 1 or a 0.  If the word were 10 bits long, there would be a total of 210 combinations. Say we've got 3 bits, and sampling rate of 1 sample per second. Three bits provides 23, or 8 quanization levels, 0 - 7. Now say we want to reproduce the level 5 during the sample. In the single-bit processor, we'd turn the thing on all the way for 5/7 of a second, and off for 2/7 of a second, so that the average level during the entire sample period would be 5. With a 3 bit processor, however, we'd have a separate amplifier for each bit, all of which would be on or off for the entire sample duration of 1 second.   The first bit would have a level 4 output, the second a level 2, and the third a level 1. To make a level five output, the first and third bit would be on, and the second one off, so that 4 plus 1 equals 5. Different ways to achieve the same thing.

Both designs have advantages and disadvantages. Single-bit designs inherently guarantee amplitude linearity, while multi-bit designs are
harder to keep linear in very high resolution applications because the bit values have to be so small, and are therefore hard to manufacture within necessary tolerances without the use of laser trimming. For instance, if your second bit accidentally were set to put out a level 2.2, then it's no longer in the correct ratio to the other bits. A single-bit only has to be proportionally correct to itself and, regardless of the value, it is. On
the other hand, single-bit designs also generate a lot more noise in the conversion process, which must be then filtered out and isolated lest it
interfere with other components in the player, such as the clock, which could introduce jitter and subsequently distortion. Is either one better?
No, just different, and the performance depends entirely upon the parts and design of the machine as a whole. The ultimate resolution is determined by the linearity, distortion, and noise floor of the system, including the analog components. For instance, many "24 bit" players actually only have 18 bits of resolution, and some 24 bit converters use single-bit technology and achieve 20 - 21 bits of resolution. Is it even important when we listen to 16 bit recordings (at least for now)? Yes, because many of the 16 bit recordings use noise shaping techniques to increase their resolution in the more audible band, so that the playback system must surpass 16 bit resolution to take advantage of that.

As for DVD players, 9 or 10 bit refers to the video converters which do essentially the same thing described above, but at higher frequencies.


Divider

Q I want to purchase a 35" to 40" direct view TV in the near future. I've narrowed the choices to three makes with prices ranging from $1,500 to $3,600. If I were to spend say $3,600 on a 40", will it be backward compatible with HDTV and a wise use of my money? My viewing habits are 90% from cable and 10% renting movies. My concern is the tradeoff in spending $3,600 now on a 40" that may not be compatible with HDTV or spend say $1,500 on a 35" figuring I don't lose as much if it's not adaptable to HDTV. I'd really like to get the 40", but I'm torn between the larger screen and the premium price it commands.

Second, in your report from the 1998 CES show, you mention the advantages of the new audio format coming out as Digital Audio Discs (DADs). You mention current DVD players can play these discs but will only resolve 20 of the 24 bits. The next generation DVD players will be able to resolve all 24 bits. Do you know what the time frame is for these DADs to hit the shelves? I'm also thinking about purchasing a Marantz 5 disc CD changer and a Sony DVD player, but I'm wondering if these are wise purchases versus waiting for a player that would play all formats such as CDs, DVDs and DADs.


A All TVs will be able to view HDTV broadcasts, but our current NTSC sets will require an adapter that converts the HDTV signal to NTSC. The 40" TV you are considering will be able to use this adapter just as any other TV will. However, the image you see on that TV will not be an HDTV image, but rather, a standard NTSC image (like we have right now). So, if you are really thinking ahead to viewing HDTV images, you will need to wait until the HDTV sets arrive in the stores. Instead of $3,600, they will probably be in the range of $5,000 - $7,000, but we won't really know the price until they actually are here. HDTVs will be able to view all DTV formats as well as NTSC. I have seen some 35" NTSC sets for around $1,000, and 32" models for $600. My suggestion is to get one of the 35" or 32" sets at as good a price as you can find, for now, and save the money for HDTV. After you get the HDTV, you can move the other TV to a different room. If it's going to be the kitchen, I would suggest the 32" or 29" set. Spouses don't usually like large TVs in the room when dinner is by candlelight! I was absolutely blown away with HDTV images at the CES, and you know how seldom I use that description.

All DVD players, including the first generation, will play CDs, DVDs, and DADs. It's the resolution capability that has changed with the latest generation. All you need to do is look at the spec sheet on your potential DVD purchase to see if it says it will decode 24 bit / 96 kHz digital bitstreams. Some of the earlier ones said 96 kHz, but not 24 bit (these are the ones that are 20 bit). It is hard to say when DADs will become available for the mass consumer. In the meantime, however, you should not be worried about going ahead with buying the DVD player, since the DADs will play on them when the discs finally come to market. There is always the possibility that some other form of DVD audio will also rear its head, but technology is never at a standstill these days. There is no assurance of total forward compatibility with any product.

Divider

Q I understand that the " .1" LFE channel in Dolby Digital carries frequency information from 3 Hz to 120 Hz. Does this mean that the other 5 channels are devoid of frequencies of 120Hz and below? Or is the LFE channel included to compliment the other channels by accentuating some of their info.? Example: Let's say a setup includes "full range" speakers for all five full frequency channels. Would a separate sub hooked up to the LFE channel be necessary?

A The LFE channel is designed for separate low frequency effects, but it is not generally used that way right now. Instead, the LFE channel carries a duplication of bass from the other channels. This makes it easy for a subwoofer hookup, but, for the time being, is a waste of the technology. At this point, the only advantage to using the LFE output from a DD or DTS decoder is that you will bypass any crossover that exists in the signal path of the regular subwoofer output of a receiver.

Divider

Q Is it possible to connect the "color stream" jacks to a RGB scart? The jacks have a red green blue color, but there is no jack for the synchronization.

A No, not without a Component-to-RGB converter. Component is a matrixed version of RGB. There are a couple of DVD players on the market that offer RGB with or instead of component.

Divider

Q My question is concerning S-Video coming from an outboard AC-3 decoder. If I were to buy an AC-3 ready A/V receiver that did not have an S-Video jack on it, could I later buy a decoder with S-Video output in order to make a full S-Video connection? (I plan on buying a TV that has one, and a DVD player in the future.) I know you can help me!

A Your A/V receiver does not need an S-Video jack to enjoy the benefits of DVD with S-Video. Just run the S-Video cable straight from your DVD player to your TV. This is not the most convenient way to go but gives the best image, even when the receiver has an S-Video connection. But, if you buy a decoder that has an S-Video switcher in it, then you can use it, and it will be especially helpful if you have several S-Video sources.

Divider

Q If laserdiscs are capable of working with DTS decoders by just hooking up a digital cable between the PCM output and a DTS decoder,
why is it such an issue for current first generation DVD players not being able to send the same bitstream out of their own PCM output (assuming the discs have that data there)? Am I correct, or am I not, that all first generation DVDs are incompatible with DTS decoders? I had read in a magazine that both the DVPS3000 and DVPS7000 Sony units were DTS capable. Do you think this was a misprint or were they referring to CDs played back on the machines using the PCM output?


A From what I have heard, DTS had originally planned on using the PCM track for DVD. The DVD consortium said NO, they were not allowed to use the PCM track. This is why both first and second generation DVD players will NOT play DTS DVDs. This includes Sony. Panasonic, Denon, Pioneer, Faroudja, Meridian, and Runco all have players on the way that will play DTS discs, but none are here yet. I asked one of the Sony technicians at CES about this subject, and he said that even their 3 new players scheduled to come out later this year will not support DTS, although this could change, and I hope it does. I hate to say this, but DTS has really dropped the ball, and actually they have been fumbling since the very beginning with laserdisc. This is unfortunate, because DTS is a great format.

Divider

Q (1)  What advantage does digital convergence have, and what is the other type (analog convergence)?

(2)  What allows certain projectors to support various aspect ratios, while others can't?


A (1)  Analog convergence uses potentiometers on the projector. These tend to vary over time at a greater degree than digital ones. It also requires you to do the adjustments from the projector. One slip while adjusting these pots and you will have to start over. With digital convergence, you can make adjustments from your remote control. The values are usually represented by numbers on the remote, so you can write all the initial values down and then start playing with it. You should always be able to get back to where you started because of the known values you started with.

(2)  It is the projector's innards that allow multiple aspect ratios, and this feature is an added cost that the budget projectors leave out to keep costs down. For every aspect ratio, there is a requirement for additional memory to store the convergence and gray scale info. A lot of the new projectors coming on the scene now include multiple ratios as standard.



� Copyright 1998 Secrets of Home Theater & High Fidelity
Return to Q&A Index.