Yes...I do. Once the gross corrections are made to the track to get them close to a proforma spectrogram and cumulative spectrum state, the back-and-forth fine tuning then occurs via listening. Typically that process takes the most time and effort, and is done at as close to concert volume for each track as is possible--subject to the limitations of the other in the household working in a different part of the house behind closed doors.
My setup and room are fairly well calibrated (details). I've also found it necessary to adjust my listening position/chair to avoid nearfield reflections within a metre or two of my ears. This is at least as important as calibrating the system. In effect, I'm using a home studio of perhaps at least as good of acoustic performance than many for-profit studios.
So the magnitude of the EQ for that example caught your attention? That was the purpose of that example. I have many more examples that are less severe.
Because of the process that I use, the psychological decision making bias effect of "anchoring" (Kahneman & Slovic) is avoided.
Chris
My setup and room are fairly well calibrated (details). I've also found it necessary to adjust my listening position/chair to avoid nearfield reflections within a metre or two of my ears. This is at least as important as calibrating the system. In effect, I'm using a home studio of perhaps at least as good of acoustic performance than many for-profit studios.
So the magnitude of the EQ for that example caught your attention? That was the purpose of that example. I have many more examples that are less severe.
Because of the process that I use, the psychological decision making bias effect of "anchoring" (Kahneman & Slovic) is avoided.
Chris
Last edited:
Sausage.
No...but disturbing to those that have been used to thinking about their music tracks as "reference".
[Earth-centricity was so firmly believed by the rank-and-file that Galileo was excommunicated until relatively recently for talking openly about heliocentrism.]
I don't recommend ignoring the reality of why our sound reproduction systems have had so much trouble reproducing something representing reality. Once the music tracks are returned to something representing real performance, I've found that new life is immediately breathed into the music again.
Scary...but true.
Chris
This is a subject I am shocked that I know nothing about, in spite of taking part (as a musician) in classical music recordings, and often going in to the box to hear playbacks. If I may a few questions.
Assuming an orchestral piece, is the conductor involved in the mastering decisions?
If you have a mastered recording, how is the unmastering curve arrived at?
Was there a time before which this doctoring did not take place?
Assuming an orchestral piece, is the conductor involved in the mastering decisions?
If you have a mastered recording, how is the unmastering curve arrived at?
Was there a time before which this doctoring did not take place?
I think that the musicians usually take a back seat when the music recordings go to the mastering houses. This seems to be the dividing line between artist control and "we'll take care of that part for you: trust us..."
Basically, the cumulative spectrum for the incoming track is examined closely, then the most fundamental EQ tilt is tried, undone, iterated, then tried again. Once the overall shape of the cumulative FR curve begins to look correct, then the local peaks and dips are addressed in the same manner. Then the track is listened to.
Any LF noise can then be identified on the spectrogram log(f) view at this point and removed by backing up to before the EQ curve is applied again then apply extremely sharp (Q=20) notch filters on those 50/60 Hz and their subharmonics (It turns out that no notes on the even tempered scale coincide within 1 Hz of 50 or 60 Hz, so the notch filters typically remove only line noise).
If the EQ is then reapplied to the track and sounds good on first try then the EQ is iteratively changed based on the listening, perhaps also going back and listening to the original again. Usually going back to the original track (un-EQed) isn't useful because the original tracks are so far away from listenable that it provides no useful comparison. However, if the EQ is relatively light overall, then the comparison with the original helps to identify the octave bands that are still requiring some adjustment.
Once the track has reached a final stage where the results sound realistic and pleasing all the way through, the track is saved. The resultant EQ curve is automatically be saved in the working buffer for potential reuse on later tracks on the album, or the curve can be saved locally or exported to XML to be used by others.
Note that at this point, only the notch filter(s) and one resultant EQ curve has been applied to the track.
At a later date, the track is listened to again, and any necessary tweaks can be made. As it turns out, once the track is in this "proforma" state, it takes very small changes in its EQ to have large effects on its sound. That is an observation that is noted to those just starting out that there really is a natural state that is both audible when found, and also highlights the heavy-handedness of the EQ that is typically applied to the tracks during their mastering before the consumer gets them. It doesn't take much EQ to get what is needed, as it turns out, if you are close to the "as recorded" state.
I've gone back to recordings made in the 1950s (mono, and stereo in the early1960s) and have found the same patterns of EQ applied to attenuate bass and accentuate HFs above 1-2 kHz, over and above the RIAA phonograph EQ curve (i.e., the -40 dB end-to-end curve). Once these recordings have been "unmastered", they typically become extremely listenable even if the sound quality has been significantly damaged due to the limited recording and mastering hardware performance/mastering practices of that time.
So to answer your question: I haven't found one.
I believe it can.
I've found 1/8dB over an octave can mean the difference between real sound and speaker sound. Linkwitz has found similar and IIRC wasn't sure how to explain how this small change can dial it in. Around 10k I've noticed the roll-off should match the expected natural attenuation for the image distance.
I had a feeling you had some Jubilees though I can't remember reading about it.
It's good to know that I'm not the only one that's found that sensitivity to EQ. I was really surprised when I stumbled onto it. I thought that something else must have been occurring in the process, so I looked for some time before accepting that the human hearing system can tell when things are about right and when they aren't.
The scary thing is...the music recording industry has consistently ignored that phenomenon and proceeded to crank in lots of EQ to make the recording have a "signature sound" or at least something that they believe somehow sounds better on really poor auto stereo systems of that early era, or at least "nearfield monitors" of the Yamaha NS-10M variety that are supposed to represent a "typical consumer stereo sound". I think you know this argument. However, I've found that the notion of EQing for a set of poor loudspeakers is fundamentally flawed.
I believe that the entire music recording industry is driven by memeplexes that are not actually correct for the typical listener today, resulting in the state of recordings that we have presently. I think this stems from the fact that no one really knows what makes hits, so bits and pieces of domain memes appear and get entrenched in the A&R industry's standard operating procedures. Consumers have been listening to music tracks continuously such that they have become acclimatized to "that sound". Musicians that play in ensembles of acoustic instrumentation come to accept the fact that music on stereos and live music never sound the same.
It doesn't have to be that way, however. The baggage of history has made the recording industry what it is today, I believe. When you look at the introduction of the CD format 33 years ago, the stories that are told of the industry's response to it are quite revealing. Nowadays, music releases don't have to be "one size fits all" since the cost of distribution has essentially gone to zero. (See The Long Tail for more discussion of the economics of recorded music distribution and profits.)
Chris
The scary thing is...the music recording industry has consistently ignored that phenomenon and proceeded to crank in lots of EQ to make the recording have a "signature sound" or at least something that they believe somehow sounds better on really poor auto stereo systems of that early era, or at least "nearfield monitors" of the Yamaha NS-10M variety that are supposed to represent a "typical consumer stereo sound". I think you know this argument. However, I've found that the notion of EQing for a set of poor loudspeakers is fundamentally flawed.
I believe that the entire music recording industry is driven by memeplexes that are not actually correct for the typical listener today, resulting in the state of recordings that we have presently. I think this stems from the fact that no one really knows what makes hits, so bits and pieces of domain memes appear and get entrenched in the A&R industry's standard operating procedures. Consumers have been listening to music tracks continuously such that they have become acclimatized to "that sound". Musicians that play in ensembles of acoustic instrumentation come to accept the fact that music on stereos and live music never sound the same.
It doesn't have to be that way, however. The baggage of history has made the recording industry what it is today, I believe. When you look at the introduction of the CD format 33 years ago, the stories that are told of the industry's response to it are quite revealing. Nowadays, music releases don't have to be "one size fits all" since the cost of distribution has essentially gone to zero. (See The Long Tail for more discussion of the economics of recorded music distribution and profits.)
Chris
Last edited:
I've found that looking at the spectrogram log(f) view is critical to getting the EQ properly set above 10 kHz. The cumulative spectrum curves typically show an unnatural roll off above point that if you try to boost back to the 1/f log-log curve will not sound very good.
In fact, the spectrogram log(f) view is critical to fine tuning across the entire track, especially if it is of long duration. Setting the view to automatically adjust its relative loudness colors to an -18 dB/decade seems to work perfectly for me.
I've owned the Jubs since late 2007. They've been the single best investment in my listening room--especially with the TAD 4002's. I've learned a lot about how to set them up with acoustic treatments to get the most out of them mostly through talking with Roy Delgado (PWK's heir apparent) and doing acoustic treatment experimenting and measuring. I wish that more people could hear them to understand how different they sound from Khorns.
That experience was the incentive for the "Corner Horn Imaging FAQ" wiki article, a version of which is found on this site. I need to go back and update that article based on new things that I've learned since it was created.
Chris
Here's to the new forum hardware and wiki software. May the wiki be back up in 7 hours
This is really interesting what you're doing Cask. This can solve a big problem with HiFi. Subjectivists like things to sound 'to taste' and objectivists want a system that does not get in the way. Now maybe we can agree on having a system that gets out of the way, but change recordings 'to taste'.
Interesting
And unrealistically optimistic.
Interesting
And unrealistically optimistic.
Last edited:
Is it? Clearly it works for me. Am I special? Probably not. That's why I shared just a small nugget of my experiences doing it.
I believe that unmastering your own recordings actually bothers those most who only want to buy everything instead of spending a little time (and gained experience) doing it. It costs nothing but a little time.
Shouldn't that be a major portion of "DIY audio"? It seems odd that it shouldn't. More than building your own hi-fi hardware, unmastering is DIY because it's at the heart of the experience: better enjoyment of the music itself.
Chris
No, I don't.
I can tell you that there is a lot less of EQing going on with multichannel (5.1) music.
One of the gifts I got for the holiday season was Beethoven's 9 symphonies on Blu-Ray (with video) of the Bavarian Radio Orchestra under the direction of Mariss Jansons (recorded 2012 in Japan). I have to tell you that the difference between the 5.0 and stereo sound tracks is so spectacular that we watched the 5th twice, back to back, when I figured out that we were watching in stereo. There was that much difference. Breathtaking.
Chris
I can tell you that there is a lot less of EQing going on with multichannel (5.1) music.
One of the gifts I got for the holiday season was Beethoven's 9 symphonies on Blu-Ray (with video) of the Bavarian Radio Orchestra under the direction of Mariss Jansons (recorded 2012 in Japan). I have to tell you that the difference between the 5.0 and stereo sound tracks is so spectacular that we watched the 5th twice, back to back, when I figured out that we were watching in stereo. There was that much difference. Breathtaking.
Chris
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.