once you oversaturate, it's lost.
^
This is quite true. The important thing is to avoid clipping (oversaturation). To do so, one needs to reduce input level. The question is how to go about reducing the input level.
Bass-rolloff (also called a high-pass filter) reduces input level in a frequency specific way. It reduces the level of low-frequencies more than the level of the midrange and treble frequencies. In that way it acts as an EQ adjustment in addition to reducing overall level, which tends to be dominated by low-frequency content.
Turning down the recorder's input gain reduces input level by the same amount across all frequencies. It reduces the bass, midrange and treble by the same amount. It does not change the EQ balance.
Both can eliminate clipping or oversaturation in the recording.
If an recording will benefit from EQ adjustment which reduces bass content, then using bass-rolloff on the recorder is one way of achieving that EQ adjustment. Another more-or-less equivalent method is switching in a high-pass filter elsewhere in the signal-chain before the recorder, such as at the microphone itself or in an external preamp if one is being used. Switching these filters in also reduces overall recording level, but that's really secondary as that reduction in level can be made up for or not by using more or less input gain. And the shape of the filtering and the frequency point below which it begins to become active varies between devices, some cut more sharply but start at a lower frequency, some the opposite. Some provide a few different low-pass filters to choose between.
The argument for not using high-pass filtering when recording is that one can achieve the same filtering after the recording has been made, or can better tailor the amount and shape of the low frequency reduction to appropriately suit the recording. The key to doing that is recording with the input gain set low enough so that the low frequency content does not clip the recorder. This approach allows for more nuanced control over the sound of the end-result, but requires doing that EQ work after the recording has been made.
An argument for using high-pass filtering is that it may reduce or eliminate the need to make an EQ adjustment afterwards. It won't do so as accurately in a tune-able way because one is stuck with whatever shape of low-pass filtering the equipment manufacturer implements. The high-pass filtering on recorders and microphones is provided primarily as a method to reduce handling and wind noise in dialog recordings, or to reduce non-musical rumble and low-frequency bleed in isolated tracks which have no meaningful low frequency content themselves. The shape of the high-pass filter in a small all-in-one recorder is designed with this in mind, and not specifically designed for reducing bass levels in an even and musical-sounding way for full-spectrum content. However, these filters can and obviously are used that way with various degrees of success. It is certainly faster and easier to implement than filtering afterwards, and that aspect should not be disregarded. But if highest quality is the more important goal, then filtering more-carefully adjusted by ear and applied afterward is more appropriate.
Different strokes for different folks.
In the end, all roads lead to Rome.
If it sounds good it is good.
I understand what you're saying, and the lo/medium/high gain on the DR-2D (100 levels of input x 3 gains = roughly 160 different settings, as I've found each setting does produce some overlap when it comes to equal input levels) really helps in that regard.
the Edirol R-09 had a mere 60 (low gain/high gain, and each could only be set 1-30 vs. 1-100), and that just didn't get the job done.
if you're taping regular rock music on back to mellow stuff, a lo-cut is most likely not necessary, as the low-end at the show is not overbearing.
but to just discount low-cuts across the board out of the gate, to me, is a recipe for disaster.....I've even done shows with the low-cut that still came out like shit (VH '98 Anchorage and any time I taped the Crue except Vegas '97)...which to me was baffling, as the peak meter on the analog deck was barely tickling -6dB, (and that was with the use of the -20dB attentuator, which I assume was a fancy word for 'low gain'...it's been so long) yet they still came out like shit....and no amount of EQ could get rid of that burpy, distorted low-end.
someone like you I'm sure I could learn a *lot* from, as your posts are informative and well thought out...it's a shame that more here aren't like you.
The point that everybody is trying to make is that the above bolded statement isn't how it works. If you set gain correctly, using an HPF is unnecessary and you can pull down the unwanted low frequencies after the fact. This will have the exact same effect as using the low cut on the recording device. The reason the shows you mention sounded like shit is probably because they sounded like shit in the room. If the low cut has a positive effect on the recording, you can have a more accurate positive effect using EQ in post processing. Always.
For example: I taped the Sword about 6 months ago. CA11 omnis > ugly battery box > Sony M10. The low freq content in the room was staggering. I could have put on the HPF on the M10, but then I would just lose that content. Instead, I elected to just pull down the overall input gain on the recorder, allowing me to safely record the show without clipping and give me the option of removing as much of the sub content as I wanted in post. If the show was going to distort at the mics, it would distort at the mics whether I use the HPF or not.
this is what i remember, from when I initially got the Sonics (before NYE '93, taped Satriani in San Jose) thru taping Metallica on the summer shed tour and Woodstock '94
---my bass levels were *horrid*
---I explained this to Leonard (Lombardo, who build the mics), and he explained to me in great detail the 3-way lo-cut he developed
---I decided to hold off initially, as I thought it could get worked out
---after the 4 Metallica shows in CA of July that year...the bass was just too much. whether it was the fault of the D6, or how he built the mics, I've no idea, but I shelled out $150+ for the 3 way lo-cut (vs. the cheaper single-pass filter he built).
and, magically, other than the previously mentioned VH and Crue shows, that problem disappeared *entirely*, even when I switched to flash-card recorders (in '08)
now, of course, there really was no 'gain' switch on a D6, but there was the 20dB attentuator (which I assume is the same thing), and that was always "on" as well...as it seemed to desensitize the signal going to the deck
I've never had luck removing the bass on a show recorded without a lo-cut in post, as once oversaturation is reached, regardless of what you do, the bass sounds 'funky'.
the '99 Sabbath show I upped (with over 400 snatches on a *reseed* on dime alone, mind you, this wasn't a never-before-heard '72 show), you can just lock your ears onto Geezer for the entire show, as the bass is *that clear*.
ridiculously good sound.
that would NEVER have happened if I'd not used the lo-cut (7th row and 4th row for the 2 shows)
now perhaps taper positioning would help in that regard (not using a lo-cut), but I don't care for that limitation....or, you *feel* bass and hear treble, so the further you are from the stage, the less the bass has a chance of distorting an otherwise good recording.