A few things here...
Firstly, the point that sticks out to me, is that this thread is branching into two discussions, and the second discussion is kinda missing the point of the first. The purpose of this poll, is to DIRECTLY a/b the Sonosax SX-R4 and the Sound Devices Mixpre 6, from raw recordings; given how rarely we can isolate JUST this one piece of gear.... And it's not to say one is better than the other, but more so to judge how much of a difference this makes, and hopefully help educate future purchasers of both products. I myself am quite surprised at how different the two recordings sound, and as I upgrade my rig over time (as per my signature right now, the Tascam DR-40 is the obvious weak point at the moment) I will use this guide to help me think about which of these two decks I should consider saving up for, for my own personal tastes and preferences in recording and playback on my system. It seems that many here are judging these two tapes on processing which has not occurred on either tape. Again, I can't stress this enough: on the two files posted in the OP of this thread, there has been NO equalization or dynamics processing on either tape.
For those of you who are really finnicky about whether or not the resampling algorithm affected the sound of one tape over the other, later today I will be manually re-sampling the 96k Sonosax down to 48k with raw low-level C code, literally just removing every other sample and creating in a new file's metadata the 48k sample rate. This induces ZERO distortion to the output file, when downsampling from 96k; if you disagree with that, I point you to Nyquist-Shannon edit: I realized, some of you might be finnicky here; there WILL be distortion to the signal, however it will all occur above 24 kHz, and even still there will be zero phase distortion as no frequency information, for example analyzing a Discrete Fourier Transform, is being added or modified to the signal, just removed. I will manually a-b the two versions of the 96k->48k via XLD and 96k->48k via c code, and if there is a noticeable difference I will re-send Noah a new file to upload. I will also reinforce here that when Noah first received the two files he himself was double-blinded, and only after extensive discussion between the two files did I reveal to him which file was which; this reveal happened for him, after he made this post - which reinforces the standards of double-blinding.
Normalization could arguably have some impact, however like I had stated when I level-matched before normalizing the two files were within 1 dB of one another as it is; as someone who extensively studies both digital audio signal processing and psychoacoustics, I can tell you with confidence that you CANNOT hear a major distinction of 1 dB RMS between two files; the discrimination your brain can perform is minimal (if it were 2.5+ dB, then I'd say we're getting into shaky waters...). I will also re-visit the normalization process later today and check the outputs of the normalized files, to insure they are indeed still level matched to within 1 dB of each other. I imagine that both Noah and Andy normalized their files before publishing, which would make the argument of the impact of me normalizing, moot - given that their normalized data sets pre-level matching were, again, within 1 dB of one another.
Charlie, you make a good point of re-sampling and dithering down to 16/44.1, to totally level the playing field. But I think that at 24/48, the playing field is still normalized, especially if I take the steps above to literally just drop every other sample manually.
NOW... with that being said, there is a second discussion occurring, which is whether or not touching up a tape in post is some form of heresy. I have to say, there are fantastic arguments presented here, from both sides. And why I'm hesitant to chime in, is that it seems the two schools of thought are starting to get a bit heated and defensive over the method that works for them.
I'm going to first begin this discussion by stating my obvious personal belief and bias on this front. I am firmly in the "post is a good thing" camp. I also come from a VERY different school of thought from most of you - my introduction to recording music, comes from working countless hours in a studio tracking, re-tracking, adjusting, re-tracking, adjusting... later/rinse/repeat. Then I'm used to being able to isolate each component of the recording to insure it all gels together - panning, compression, EQ. Then it's all about dressing it up using mixing experience and my own two ears: delay, reverb, etc. Then it's also about creative "mis-use" of effects such as pitch shifting, odd panning, modulation, more delay, more reverb, etc. to get the desired end result as a crafted, sculpted sonic event.
Obviously, we don't have this luxury when taping; it's an entirely different set of variables in the hand we are dealt. And the primary reason why I originally became a taper, was to learn more about how we do things in this world. It's made my own studio recordings infinitely better, as I learn more tips and tricks and intricacies of the entire live taping system. As a studio guy, I've learned LOADS from other tapers that I wouldn't have thought about or tried in the studio, because we as tapers deal with certain specific problems every time we make a tape, and none of those problems are present in the studio. As a result, when I go into the studio now I have infinitely more tools and secret weapons at my disposal, that make me a better producer and mix engineer. Likewise, I think to ignore the body of work and experience that studio guys have, is only harming you in the long run; to ignore what studio guys do day-in and day-out, for a living, is IMO short-sighted. And I don't intend for that to be a harsh attack, but more of a "hmm, maybe you should at least consider what people doing this day-in, day-out have to say about how to make something already great, even better."
Look, I get those of you who say that a good set of microphones should be more than sufficient to properly capture a recording. But many of the recording systems we set up and the tapes we produce, are already inherently flawed and imperfect at the system level. By this same logic, I could argue that the ONLY way to truly and properly capture the experience of being at a show, would be to put binaural mics in your ears, or to get an (unnecessarily expensive and HIGHLY calibrated) binaural manikin and make your recording that way. Likewise, I'd argue that your playback systems are inherently flawed. And so my initial counter-argument, really, is that the entire system is always flawed, based on what approach you take. We can (and frequently) DO take steps to minimize these flaws, and in the long run we can get pretty darned close the more we work together at this.
I imagine many of you who are "no post" purists, hold your opinion partially based on either not-so-great work from others ruining recordings via "re-mastering" or have yourselves been frustrated by trying to play with an EQ, and every time just going back to the original un-EQ'd recording. Both of these are totally valid, and in the long run I'd agree with you that I prefer an untouched tape, warts and all, to a poorly-EQ'd tape. But a properly EQ'd tape? It makes a WORLD of difference. What may be worth my time here, is to perform a similar experiment to what Noah's done here - I will post two tapes: one un-touched, and one where I apply EQ as I would approach it as a studio track, and YOU can decide whether or not you want to forego EQ, not letting me tell you that you SHOULD or SHOULD NOT do it, because as usual people on both sides of this argument should be taking what I have to say with a grain of salt. RULE NUMBER ONE of audio, is to trust your ears. I can't tell you what sounds good or bad, but YOU can tell yourself what sounds good or bad.
Re: mixing to a room... I'll happily yield to that point. To which my counter-argument is, there's a reason I only mix either in HIGHLY HIHGLY HIGHLY controlled environments, or on headphones that I know and trust intimately from working on them for YEARS. My mixing environment right now is a Focusrite Scarlett or Steinberg UR824 (I plan on moving ASAP to a 32-bit/192k conversion system with decent mic pres, but that's another discussion for another time), through either two or five Genelec 8030As with a Genelec 7030B sub, customizing the crossover network, in a HIGHLY treated room that's basically totally flat, with every wall and the ceiling coated with Owens-Corning 703 and with bass traps in the corners. What I am hearing, is basically the raw monitors, set up carefully to insure maximum linearity of my system and to insure the room I'm working with has no combing. I think that's a slightly different argument than mixing to, say, my bedroom with my two B&W DM302s and highly reflective (if not diffuse - brick and old ornamented wood) room. The former, I say hell yes let me mix in there because I have both controlled for the benefits of and understood the limitations of, that specific mixing environment. The latter, I say yeah you're gonna have an insanely tough time and it's best to leave the tape untouched at that point in time.
At the end of the day, we all have the same goal: to as closely as possible capture the magic happening on stage, and re-produce that magic as closely as possible for both ourselves and for the community at large. It's a massive (and often un-thanked) service we provide to the live music community, and we suffer through all of the hardships and technical difficulties of producing a good tape (for which, I've found, the biggest threat is rando intoxicated dude/dudette that wants to yell and cause trouble and harm to our otherwise mostly peaceful section); because at the end of the day, once we hit the play button after the show, it's all worth it.