Dynaudio Impressions

I finally got around giving the Brian MR session a listen at home and again on the Dynaudios. For the purposes of clarification, “home” means a Musical Fidelity Elektra through an NAD 317 and a pair of Origin Live bookshelf models.

First, I’ll talk about mixing with the Dynaudios. Apparently these monitors cost roughly ten times the ticket price for the Samsons. Andy wondered aloud to me whether the Dynaudios could justify the price tag. I can hear the effects of the dollar, dollar bills, and these effects…
might justify the price for someone like a movie soundtrack engineer. For me, the technical strengths of the Dynaudios probably make them less desirable for mixing rock than the more modest Samsons. The Dynaudios separate frequencies and accurately reproduce frequencies from about 80 hertz to the top of their range better than any speakers I’ve yet heard. I noticed a lot more to criticize while line checking the drums than usual, and the Dynaudios helped me fine tune the sound I wanted from the kick and the snare with use of mic placement. As per usual, the floor tom sounded terrible. I couldn’t get anything with subtle mic placement, so Kyle advised me to back the mic off the head by more than a foot. I left the channel in the mix, but it seems to have had little effect. Going through a quick level check, I felt a bit distracted by the subwoofer. I don’t spend much energy trying to shape the sound below 60 hertz, so, although the sub makes me feel engaged with the music as music, it takes a bit of my attention from the frequency ranges I can control and makes me feel less engaged with the music as a maleable entity I need to polish and control. I set the low pass for the sub at about 60 hertz and turned the sub-specific volume to a barely noticeable, but noticeable level. I had to give my testicles something to do while I worked. Unfortunately, I left the high pass on both the sub and the sattelites at 80 hertz. Finishing with the levels, the frequency separation and frequency specific accuracy played their proper roles again, but this time with a downside. (I don’t know whether these terms makes any sense to audiophiles, but I mean to express the degree to which I can separate in my head sounds of different pitches and similarly pitched sounds from the same instruments.) If a guitar got a bit loud in the mix, the Dynaudios seemed to make the discrepancy between the guitar’s current level and it’s natural level seem somehow more dramatic than it should. The monitors, in other words, seemed to hold a magnifying glass to errors in the mix. In most cases, I would appreciate this characteristic, but I got into trouble with the Dynaudios. Sometimes I like to highlight particular mic channels by letting them ride a bit above the mix. I tried to highlight the kick drum with Brian MR. I had a lot of trouble finding a subtle way to put the kick out in front of the mix with the Dynaudios. When I tried tried to place it only slightly above the snare, it sounded dramatically in front of the snare. I tried experimenting by throwing the fader on the kick mic ten decibels higher. It sounded a little louder, but it didn’t sound any more fundamentally above the mix than it had two seconds and 10 decibels before. Not only did this aspect of the monitors’ performance hinder my ability to make a nice initial mix, but, throughout the night, as instrument level changes would prompt me to remix, I struggled determining how radically to remix. When it came to effects, the Dynaudios let me hear the slightest delay, but, again, the slightest delay seemed a little too wet to me. The Dynaudios held a magnifying glass to effects just as they did with levels. Just as with levels, very badly over-affected channels didn’t sound as over-affected as I expected them too given how affected only slightly wet channels sounded. In the end, my mix pushed the kick drum a little too far above the rest of the mix and the effects came out pretty well. I don’t have much confidence I could get a significantly better sound from the kick if I could mix the set again, and I wouldn’t expect the effects necessarily to sound as good as they do. Perhaps if I mixed five sets with the Dynaudios I’d have a better sense of their subtleties and more confidence. Still, not only do these monitors seem to exaggerate delay effects and discrepancies between good and bad levels for certain frequencies within a mix (I want monitors to do these things as such monitors will help me to notice and correct mistakes in my mixes), but they do so in such a way as to give me no impression how ordinary stereos will react. All good monitors require engineers to know how to translate between sound through the monitors and sound through ordinary stereos. Even with monitors very different from ordinary stereos, engineers might have an easy time translating. I feel even if I knew the Dynaudios well, the monitors would make translating a difficult task for me.

Upon listening at home, I noticed only one property of the Dynaudios I would classify as a peculiarity. A guitar I panned to the right channel tended to feed back a little at a frequency somewhere around 300-400 hertz. On my home stereo, the feed back annoyed the hell out of me. On a pair of Grado SR 60’s, the guitar sounded a little more natural, the effect of a live set, but still a little amazingly strident. On a second listend with the Dynaudios I noticed the frequency, but, somehow, I didn’t mind it. Also, as I’ve mentioned, the kick drum sound came out in front of my mix a little too much when I listened at home. The Dynaudios make the kick sound pretty natural. As I’ve mentioned, the Dynaudios seem to separate sounds very well, and the kick drum seemed no more separate from the mix than any other sound on these monitors. I listened again with the high pass switch on the sattelites set to 60 hertz. This time, the 300-400 hertz frequency of the guitar seemed a little more annoying, as it had at home, and the kick sound did seem to ride a little above the mix. I can understand the newfound agression of the kick sound, but that switch shouldn’t have affected the guitar frequency unless the high pass filter on the sattelites has a regretably shallow roll off. I admit the subtle differences I hear might have more to do with psychology than sound, but I hear what I hear. Stereo image has a lot to do with speaker placement, and we don’t have much control over speaker placement in our control room. Still, I don’t have any complaint with the image of the Dynaudios. The stereo image I wanted translated pretty well to my stereo at home.

As instruments, the Dynaudios are very precise. Some tasks require or at least benefit from very precise tools, and, for those tasks, engineers should invest time in learning about the peculiarities of their precise tools. Mixing rock music requires more precision than most people seem to believe, but I have some reason to think the Dynaudios too precise for the task at hand. If I learned to use them properly, my mixes would sound better than the Brian MR mix. On the other hand, I think my mixes would sound better with the Samsons than with the Dynaudios even if I had a profound knowledge of both. The Dynaudios seem tuned for making slightly finer adjustments than those I tend to make while mixing rock. The Samsons, though more blunt, seem better suited.

3 thoughts on “Dynaudio Impressions

  1. I’ll start by saying that I mostly agree with Nick; the Dynaudios are extremely, scarily clear, and in almost all circumstances much more clear than anything people are likely to hear from our OTA signal.

    The point we differ on is that I don’t see that as a bad thing at all. In terms of sound reproduction, we’re obviously never going to be able to replicate what Pipeline sounds like after it’s passed through the compellor, the airwaves (interweb, whatever), and some random listener’s stereo. While I have some other ideas about that (we should run a line with the OTA signal into C-control and attach a decent pair of headphones), I don’t think accurate reproduction of the final sound is a realistic goal with studio monitors. The thing I like about the Dynaudios is that they basically reproduce the gross spectral shape of the OTA signal (i.e. no weak bass, hyped mids, or anything like that) while sounding orders of magnitude better in every other respect (namely the frequency separation). This means that problems with the mix are greatly magnified in the control room; you’re going to hear every bleed, every bad mic placement, and every poorly-compressed instrument/voice in sharp relief. While it’s tempting to overcompensate for slightly exaggerated mix problems sometimes (as I think Nick mentioned), in my experience (admittedly minimal; I’ve only done three bands on them and one was so bad I didn’t bother to listen to the recording later) nailing down an adequate-to-good mix in the control room means you get a similar, but even better-sounding, final recording. The Charms and Les Breastfeeders sets are two of the best-sounding mixes I can remember doing, and I hadn’t even really started playing with the subwoofer at that point.

    The point I will agree with Nick on is that their clarity makes adding subtle effects (mostly the slight reverb that’s almost always needed on vocals) kind of tricky, as a certain amount of frequency-blurring is critical to getting the sound you want and the Dynaudios refuse to deliver it. The low-frequency feedback thing is weird, but I haven’t noticed it (I also haven’t looked for it; I will now). Might just be a design flaw in our particular units.

    Nick’s final conclusion is that the type of recording we’re doing doesn’t need the degree of precision the Dynaudios offer. I agree with him on that in general, but would argue that having such precision for our live recording isn’t a big enough advantage to justify the insane price tag of these things. To put it another way, the Dynaudios are the best monitors I’ve had the privilege of working with, but they definitely aren’t 10X better than the Samsons (which deliver almost the same clarity but have a slightly weird frequency response) even though they cost about 10X as much.

    For the record, I tweaked the subwoofer like this:

    – set the high-pass on the mains to 80Hz. It’s worth noting that you can leave the high-pass on “flat” and the mains will do pretty well down to below 60Hz without even needing the subwoofer, which is a nice quick-and-dirty feature if you’re lazy and/or pressed for time.

    – Put a fairly bassy CD that you’re reasonably familiar with (I think I was using Dalek or something) in the player.

    – Crank the subwoofer up really high, much higher than you’d normally want it.

    – Scan the high-pass knob on the subwoofer output slowly through its range (50-150Hz). Basically you’re looking for the sweet spot where there’s minimal overlap between the output of the mains and the subwoofer, but also no spectral gap. When I tried this it seemed to be somewhere between 90-100Hz, which is reasonable. This would probably work much better with a low-frequency tone generator, but I was just sort of screwing around.

    – Set the sub volume to a level that sounds reasonable. This will obviously vary depending on the user (my lower-register hearing isn’t great, so I probably use it slightly higher than most people), but the 9 o’clock range on the knob seems to be pretty good for our studio.

    This was a highly unscientific method that I pretty much pulled out of my ass on the spot, but I think it worked pretty decently. As I mentioned, it could be done much more precisely/quantitatively with a tone generator and/or real-time analyzer, which is something we should probably look into trying.

  2. To add to your woofer tweaks, I would also suggest everyone makes sure the high pass filter on the woofer matches the high pass filter on the sattelites. I THINK mismatched high pass filters might have caused my weird low-frequency feedback issue.

    Just to reiterate, I have no problem with the Dynaudios sounding either different from average speakers or more clear than average speakers. Monitors should sound different and more clear than common stereo speakers. My problem with the Dynaudios comes from the specific ways in which they differ from average speakers and the particular sense in which they provide a clearer signal.

    I definitely appreciate the reliable frequency response of the Dynaudios. Even without the sub (especially without the sub, actually), the Dynaudios allow you to forget about whether the signal going to the listener has too much of one or another kind of frequency. On the other hand, for the purposes of mixing, I don’t really mind if a set of monitors exhibits weak bass response. I can make the necessary mental adjustments pretty easily.

    Bryan suggests “Nailing down an adequate-to-good mix in the control room [with Dynaudios] means you get a similar, but even better-sounding, final recording.” I believe the door can swing both ways. Certainly, a system like the Dynaudios allows you to find and fix problems the Samsons never could. As Bryan notes, though, the very strengths of the Dynaudios create a blind spot in terms of how well frequencies and similarly pitched timbres blend together. Any blind spot can result in a problem of translation. One could claim, “If you mix with the shittiest monitors availible and manage good sound through those monitors, your mix will sound even better on a decent stereo.” This statement is false in general, but mixing with shitty monitors will make certain aspects of the mixing process easier. All monitors will create blind spots… choosing monitors well depends in part on finding the blind spots you can more easily live with, and the intended purpose of the monitors pretty well determines how well one can tolerate any given blind spot.

    No doubt, the Dynaudios are of a higher quality than any speakers I’ve ever used. Ideally, I’d like to have three sets of monitors with different strengths and weaknesses and to switch from one to the other. Admitedly, we can almost make that situation a reality by using headphones and the station’s many monitors, but not quite. For me the Samsons have enough (though not nearly all) of the Dynaudios’ strenghts with respect to our purposes and fewer critical weaknesses.

  3. NICK IS A FAG!!@!1

    Now that I’ve lowered the level of discourse here a bit, I’d like to just state that we basically seem to agree on these things beyond some minor, mostly philosophical points. The clarity of the Dynaudios definitely does create “blind spots” (defined here as discrepancies between the monitored mix and the OTA mix), they just happen to be blind spots that (to paraphrase Nick) I’ve found to be pretty easy to live with. I’ll agree that it’s a little tempting to try to “muddy up” a mix of a garage-rock band that sounds unnaturally clear and precise, but once you get comfortable with the idea of all the post-mix factors handling that for you, I like having the extra sonic data (as it were) in the air to work with; the increased subtlety it lends to the mixing engineers’ toolbox is more than worth the tradeoff of slightly mis-reproduced (is that even a word?) sound. Again, that’s pretty much a subjective thing; in general I’d always rather have too much information than too little (as would be the case with the “shittiest monitors available” scenario). I think having headphones with the OTA signal in C-control combined with these would be an excellent compromise solution, since (again, as Nick said) no monitor is perfect and it’s nice to have as many points of reference as possible (this would give the engineer the option of hearing the board signal or OTA signal through headphones, plus the monitors, all without getting up off his/her lazy ass).

    And yes, I agree, we should just get the damn Samsons. The price/performance ratio on those things was unbelievable, and their only real flaw (weak bass) is something any sound engineer can work around with minimal practice. Not that we really ever had the option of keeping the Dynaudios anyway, since they cost almost as much as the entire PA.

Leave a Reply

Your email address will not be published. Required fields are marked *