I have a real-world test case for this exact situation. My three-year-old removed both DVD and Blu-ray from the same case and proceeded to “polish” the hardwood floors with them. While one test does not make a water-tight case, for us the blu-ray continues to play flawlessly while the DVD skips/stutters/freezes like crazy on the same player. Take it as you will. She’s right-handed but I don’t know which disc was in each hand We may be getting off topic!
As a classical engineer, I have made a few videos for artists (or at least matched audio to video and re-rendered with a bit of color correction) and in each case they wanted the exact same audio master presented. One client asked for same loudness and the other wanted -23 LUFS in case it ended up being broadcast (actually we used the BBC PPM max of 6). Given I’m normally aiming for c. -20 LUFS, this doesn’t present any major issues with zero need to change dynamic range.
I enjoy your side notes. I probably agree with you about the microphones and what frequency range is important to human hearing. The other thing to bear in mind that is extremely relevant to classical recording is the selection of microphones that have excellent off-axis response. Any old microphone can pick up good signal from the front but only the best microphones (I hope this includes some mid-range options) have equally good off-axis response. ORTF wouldn’t be nearly as effective if the off-axis response was rubbish. Think of it this way: When I set up at typical ORTF distances where the choir or orchestra fills the whole of the SRA, none of the direct sound is hitting the capsules head on.
That may be a large part of what I like about the ribbon mic. I don’t know. I do record with greater mic spacing than when dealing with live sources. So the ribbon at 18 inches sounds more open than the sm58 at 2 inches on stage for vocal work even though the sm58 picks up higher frequencies. (no I don’t have any sm58s at home to compare in a recording environment)
I use ESI-nEar08. Maybe not the best but I think they are neutral enough.
How can we calibrate them properly? How to generate pink noise in Ardour?
Any good tutorial?
Looking the plugins by autor I have found the Robin Gareus “Test signal generator”. You can generate pink noise with this plugin.
Many plugins can be found only by “…autor” or “category”, all the plugins are not displayed at the “Plugin manager”…
I have no experience with these but the SOS review was positive. Detail is good but not sure how Paul White’s “a little forward-sounding” comment would work for classical. If you recognize the intricacies of your flute-playing on playback that’s probably a good thing
As per that SOS article I posted a while back #187. Use the x42 noise generator or some pre-recorded pink noise WAV at -20 dBFS and use a SPL meter to measure output from each speaker separately. SPL level will depend on room size so that you can keep your hearing for a little while longer.
Oh, yes. Sorry, I forget this article.
Well I have calibrated my speakers following the indications however now I have to turn down the volume of my system again because it is quite high.
What is the purpose then?
I did the SPL level at 80db, my room is very small and the speakers are wide separated but very near in my table. Maybe should be less than 80dbs?
I missed something…
I missed the end of the article, sorry.
I used the app “DB ABCD” in my iPad, c-weighted/slow.
The iPad was in my music stand at the position of my head.
I also did the equilateral triangle.
74dBs is much better, however is still a bit loud, in my opinion. For chamber music is ok but for jazz or pop is still a bit loud, I think.
The reason it feels too loud is probably because those jazz and pop records you are listening back to were mastered louder than the K-20 or -20 LUFS value you calibrated to. The key thing is once you have calibrated your speakers, make sure the volume of your own classical masters is at the right level according to your ears then towards the end of the mastering process check against a LUFS meter and adjust as necessary. If you find yourself having to make consistent alterations each project, you might consider switching your calibration SPL. As an example, if your masters are consistently 2db too quiet on the meter, turn your monitors down 2dB using your ipad app. Sounds counter intuitive but that will, in fact, force you to raise your master by that amount.
Remember that the latest YouTube streaming is limited to -14 LUFS (int) vs the -20 LUFS (int) that I recommended for our classical workflow. Even then, many pop/rock CDs are still mastered much higher than this (just because there are no loudness rules and because we still associate “louder” with “better”). I’m assuming you ripped some of your own CDs to test the calibration? Run some LUFS measurements and you’ll no doubt find they are louder (or much louder) than the -20 LUFS integrated you are calibrated to. Any louder than -14 LUFS they’d be turned down by the various streaming services, thereby penalizing the song two-fold by volume reduction and uncovering the squashed dynamics needed to reach that higher LUFS level. They will begin to sound quite weak against more dynamic masters.
Yes I tested with ripped CDs.
The only strange thing for me doing this process is that the indication of the output signal in the Ardour master channel of the sine-wave tone signal at -20dBs, as well as the pink-noise at -20dBs is 0 dBs, not -20dBs.
If you are using a K-20 meter on the master bus, 0 would be the correct read-out as it gives you 20 dB of headroom above that. Average values around this 0 point is what will give you lots of headroom for classical music. Average peaks to the top of the yellow portion (+4) will equate to fortissimo passages.
For my harpsichord albums so far I’m using the -20 LUFS idea but using -16 LUFS (momentary) as my equivalent +4 value. Momentary doesn’t equate perfectly to RMS but it is close enough for my needs. Note that the k-system v2 advocates for a reduced yellow zone of +3 dB. In any case, for the next album I’ll experiment with aligning the whole album to -20 LUFS (int) and see where the rest of the values end up.
The Mozart Concerto a little down the page is great for practicing multi-track mixing/mastering in Ardour or Mixbus. There is an accompanying discussion with various attempts here: http://discussion.cambridge-mt.com/forumdisplay.php?fid=426. There are other good ones too including a baroque overture, opera extracts, Bach choral and modern classical. All worthwhile to dig into!
It’s part of the online resources for Mike Senior’s Mixing Secrets Book.
Not free (either as beer or speech) but massively discounted, I can heartily recommend the Metric Halo plugins ChannelStrip and HaloVerb for classical needs in Ardour/Mixbus.on the Windows or Mac platforms. The MIO compressor is lovely on most things I’ve used it for and the HaloVerb is directly taken from the hardware DSP (I think). It creates a crackin’ 'verb for harpsichord and organ stuff. I’ve heard it work wonders on an acoustic guitar track too. I haven’t compared it too closely to PhoenixVerb, IRCAM Verb Session or Gverb+, Dragonfly on the Linux side etc but this is another serious algorithmic reverb perfect for classical audio. The sale goes until the end of the year.
Also released yesterday is Organteq from Modartt. It is Linux-friendly like Pianoteq! I think Grandorgue samples still beat it for sound at present but at 15MB install it is excellent (and will no doubt continue improving as the product matures if Pianoteq is anything to go by).
Thanks for those recommendations! I’m curious to hear what you, as a “real” harpsichord player, think of the harpsichords in Pianoteq. I started learning keyboard earlier this year to accompany my partner, a traditional singer from Brittany, and the harpsichord goes quite well with her voice (as does the organ; I’ve been using Organteq alpha for several months now and decided to go for the official release now that it’s out).