Frequently Asked Questions
Why choose Opimod-FM 8400?
8400 has been winning shootout after shootout when the criterion is pure sound quality, because no other processor gives you 8400's source-to-source consistency, loudness, low speech distortion, ease-of-use, and bright, silky-clean high frequencies. It's processing taken to the next level.
I've heard that the 8400 has so much delay that you have to use a special headphone feed for live talent. Are you ever going to make it easier to install the 8400 in a live-voice environment?
Your wish is our command. Software version 2.0 (available in the last part of February, 2002) cuts the 8400's delay in half (to about 20 milliseconds), which is comfortable for talent because they no longer hear echoes in their headphones. You can now drop in the 8400 in place of an older processor with no studio wiring changes. And off-air cueing of remote talent is no longer a challenge.
How will the 8400 accommodate IBOC digital systems?
Starting in summer 2002, we will offer an "IBOC Output" option on 8400, which can be retrofitted into all existing 8400s at modest cost. This additional AES/EBU output carries the output of the multiband compressor/limiter as fed directly into an advanced look-ahead limiter. By tightly controlling peak levels without clipping, this output prevents the IBOC codec from wasting bits by encoding clipper-induced spectrum, yielding the finest possible sound at a given bit rate. You get no-compromise audio quality on both the analog and digital channels simultaneously.
Who is behind Optimod-FM 8400?
A terrific team of hardware, DSP, and control system engineers, headed by Bob Orban (Vice President and Chief Engineer) and Greg Ogonowski (Vice President, New Product Development). Between them, Bob and Greg have 50 years of audio processor design experience. This award-winning team combines advanced academic degrees and theoretical DSP knowledge with very recent major-market radio engineering experience. Both Bob and Greg have legendary ears, and the renowned Gregg Laboratories processors that Greg created are still some of the most sought-after processors on the used market. This combination of advanced theoretical knowledge, real-world experience, and competitive ears is unmatched by any other processor company.
How can you remote-control 8400?
8400 comes with software that uses the Internet TCP/IP protocol to communicate with 8400 either via its serial port (for modem or direct cable connections) or through Ethernet, via a PC card that plugs into the 8400's rear panel.
Can you automate the 8400 by time-of-day?
8400 has a built-in real-time clock. You can program many different events (like recalling a given processing preset) to occur at preset times.
Does 8400 have a composite clipper?
It has something better—a patent-pending composite limiter that is as loud as a composite clipper but which protects the stereo pilot tone and subcarriers (including RDS) far better. However, because of its excellent left/right-domain peak control, 8400 doesn't need composite limiting to get loud, so you can locate the 8400 at the studio side of an uncompressed AES/EBU digital STL while still achieving high on-air loudness.
Is 8400 compatible with 32 kHz sample rate uncompressed digital STLs?
Completely. Unlike competitive products that use unnecessarily high internal sample rates, an uncompressed 32 kHz link adds no overshoot to the 8400's AES/EBU output signal.
What input sample rates does the 8400's AES/EBU input support?
Anything from 32 kHz to 96 kHz works fine.
Why does the 8400 have more input/output delay than other processors?
Because the 8400 uses "look-ahead" techniques. These techniques delay the audio in the processor so that its control software "knows" about audio waveforms that it has not yet processed. Compared to processors that don't use look-ahead techniques, the 8400 can use this knowledge to make better decisions about how to adapt the processing to the program material as it flows through the processor. At this time, the 8400 is the only available processor that can create state-of-the-art loudness on music and extremely clean live voice within the same preset. Low-delay processors that emphasize music loudness invariably have more voice distortion than 8400 (particularly on male voice), which is sometimes so blatant that it makes the mic channel sound broken. The 8400's look-ahead techniques also reduce music distortion without compromising loudness.
(Historical Note: Although one competitor's website has the gall to claim that they "invented" look-ahead limiting for fine-arts formats, the actual inventors were Shorter, Manson, and Stebbings at the BBC's Research Department, who first developed practical on-air look-ahead limiting in 1967—when said competitor's founder was still in high school. The 8400 uses various novel forms of look-ahead processing in several different places within its DSP algorithms. While inspired by the original BBC research, Orban's techniques are original and represent a considerable advance over the BBC's pioneering work. They work great for the most competitively processed formats, not just "purist" formats like classical and jazz.)
How did you succeed in reducing the time delay in version 2 of the 8400's DSP software?
Delay in version 1 came from two separate sources: irreducible delay caused by the 8400's sophisticated look-ahead algorithms, and reducible delay caused by various "housekeeping" functions like filter delays. By rigorously examining every reducible part of the delay, we were able to greatly reduce it without compromising the operation of the look-ahead algorithms that give the 8400 its uniquely favorable loudness/distortion tradeoff. Some of the techniques we used included parallelizing certain algorithms and running more of the processing at higher sample rates.
I've heard that 96 kHz sample rate gives you "more resolution in midrange frequencies" than 32 kHz. If so, why does the 8400 use a 32 kHz sample rate?
Actually, the 8400 is a "multirate" processor that uses whatever sample rate is right for the job at hand. Sample rates range from 32 kHz to 512 kHz, with all audio clippers running at 256 kHz.
The resolution in a given frequency band is nothing more than the noise floor in that band. The number of bits in each digital word determines this—the sampling rate has nothing to do with it. The 8400 uses 24-bit processing throughout, with the equivalent of 48-bit processing in some critical circuits. The resulting resolution is always better than the source material.
In fact, 32 kHz ideally complements the physics of the FM stereo system. The 57 kHz RDS (RBDS) subcarrier has significant energy down to 54.5 kHz, which means that protecting it requires limiting all significant audio energy to 16.5 kHz and below. By limiting energy to 16 kHz (the Nyquist frequency for 32 kHz sampling), the 8400's basic sample rate gives you a 500Hz guardband. Yet the 8400's basic frequency response is still flat to 15 kHz, thanks to sophisticated systems design throughout the processing.
In contrast, 96 kHz systems waste bandwidth and DSP resources without benefits to anyone but marketing departments Because their intrinsic 48kHz audio bandwidth cannot be accommodated by the FM stereo system, 96 kHz systems need expensive and redundant filtering to prevent aliasing distortion between the stereo main and subchannels (and vice-versa), and to prevent interference to the stereo pilot tone and RDS.
Indeed, even if 96 kHz did yield increased resolution (which it doesn't), the necessary filtering to accommodate the signal to FM would eliminate the "extra" resolution anyway. Further, the filtering in these systems is often not good enough to protect the RDS subcarrier.
Whoever told you that higher sample rates yield more resolution than lower sample rates simply doesn't understand the classic mathematical theories of sampling and information. These were first published by Nyquist and Shannon more than 50 years ago and have withstood every scientific and engineering challenge since then.
What about "compressor aliasing" at 32 kHz sample rate?
You may have also heard that operating a compressor at 32 kHz can cause "aliasing." This is true for some naïve digital compression algorithms but not for the ones used in the 8400. As usual, the devil's in the details.
Why does the 8400 apply pre-emphasis before the multiband compressor/limiter stage?
To prevent distortion at high frequencies. Many people have described the high-frequency texture of processors that don't do this as "gritty and harsh." That's because such processors have no way of accurately compensating for pre-emphasis-induced clipper overload. In contrast, the 8400's limiting system is pre-emphasis aware, and prevents excessive upper-midrange and high frequency clipping.
If you look at the details, you will see that the 8400 compressor/limiter is actually only aware of pre-emphasis up to 6 kHz. A special high frequency limiter (which uses a moment-by-moment analysis of the distortion caused by the final clipper) controls the high frequency drive to the final clipper between 6 kHz and 15 kHz by adding gain reduction as necessary to the highest band in the multiband compressor/limiter. The result of this sophisticated, split-band system? The 8400's high frequencies are bright, yet silky-clean.
Why does 8400 combine its multiband compressors and limiters in one stage?
Because that way the two can "talk" to each other intelligently. In particular, the compressor can react to what the limiter is doing, which is impossible in systems where the limiters follow the compressors. Starting with version 1.0, 8400 now allows you to adjust the attack time of the limiters as well as the attack time of the compressors, giving you more versatility in adjusting the dynamics of the sound while still preserving the advantages of the coupled topology.
When a salesman nitpicks certain minutiae of sound quality, like "flattened snare drum hits," what's his real agenda?
Magicians call it misdirection: "Listen to the snare drum; pay no attention to the egregious voice distortion or the harsh highs behind the curtain." The salesman wants you to concentrate on unimportant details to prevent you from listening carefully to pure speech, which is a hugely important part of any radio format, and which the 8400 handles dramatically more cleanly than its competition. Or he might want to distract you from focusing on the overall cleanliness of the high frequencies.
The Germans have a word for the most important aspect of the 8400's sound: its Gestalt. Roughly speaking, it means "the big picture" or "the whole enchilada." 8400 has the unique ability to create a clean, consistent, punchy presentation that complements and facilitates the flow of the various elements in your broadcast. It's something you really only appreciate after listening for a while. And it's something that 8400 does better than any other processor. 8400 processing is as close as anyone's yet gotten to "The Perfect Sound."
Back to Optimod-FM 8400 Signature Series Overview