Stanford and PEQdb Hearing Seminar by Saurav Chala (aka Sharur)

[totti1965]

Saurav Chala, yes the sharur guy who was banned from this forum two years ago because he was simply too rude and impolite made a seminar about headphone and IEM listening.

He is discussing the research he made with his company PEQdb for improving the Harman 2018 target curve.

From 1:04:57 min on he is doing some concrete suggestions for using the PEQdb site for concrete listening improvements by equalization.

I think, he did a good job in his research by incorporating a big number of volunteer listeners for finding better target curves:

 

[Blockader]

A very controversial name in audio circles. Coming from a culture where ego, toxicity and competitiveness often run high, I can look past the aggression and focus on his intentions. I’m glad he's now putting his energy and ambition toward something more constructive.

 

[totti1965]

Coming from a culture where ego...... often run high......

I think it is not the culture, its simply... S H A R U R!

 

[InfiniteJester]

I tried his tool, which I found interesting, but the results weren't good for my personal listening experience. I gave it several tries, but the adjustments were never of my liking, and the target curve they developed doesn't suit me at all. I remember that the measurements from several of the headphones also looked very strange.

I would like to read other people's experience with it.

 

[Curupira]

I tried his tool, which I found interesting, but the results weren't good for my personal listening experience. I gave it several tries, but the adjustments were never of my liking, and the target curve they developed doesn't suit me at all. I remember that the measurements from several of the headphones also looked very strange.

I would like to read other people's experience with it.

Well, I think he's on to something. On my second try, I noticed the sliders don't exactly do what you intuitively would expect them to do, so I stopped following my preconceived notion of what my IEMs EQ curve and sound should be (what I usually EQ to), and started actually listening to the songs, trying to find what would strike me as pleasant sounding.

My IEM of choice is the Letshuoer Mystic 8, which is definitely overpriced in today's market, but whose quality and comfort stops me from selling it short:

My preferred EQ Curve for it, which I came to after extensive testing and biased adjustments using squig.link (I seldom use AutoEQ, and I know my ear modes by heart),looks like this on Peace

And here is PEQdB's suggested 15 band curve:

Below, what would both curves would be following squig.link's predictions (red being PEQdb's/Sharur's)

 

[totti1965]

Well, I think he's on to something. On my second try, I noticed the sliders don't exactly do what you intuitively would expect them to do, so I stopped following my preconceived notion of what my IEMs EQ curve and sound should be (what I usually EQ to), and started actually listening to the songs, trying to find what would strike me as pleasant sounding.

My IEM of choice is the Letshuoer Mystic 8, which is definitely overpriced in today's market, but whose quality and comfort stops me from selling it short:


View attachment 487559

My preferred EQ Curve for it, which I came to after extensive testing and biased adjustments using squig.link (I seldom use AutoEQ, and I know my ear modes by heart),looks like this on Peace
View attachment 487574

And here is PEQdB's suggested 15 band curve:


View attachment 487564


Below, what would both curves would be following squig.link's predictions (red being PEQdb's/Sharur's)

View attachment 487571

… seems quite close.

 

[Eddie_Vanjovi]

A very controversial name in audio circles. Coming from a culture where ego, toxicity and competitiveness often run high, I can look past the aggression and focus on his intentions. I’m glad he's now putting his energy and ambition toward something more constructive.

I'd like to think his passion for audio would have some positive contribution to the audio community, but I fail to see what he has actually added value to. The Harman Reference Curve has already accounted for Sharur's preference in his original paper. Harman discovered variations in preference in both the treble and bass shelfs, which explicitly demonstrates Sharur's lack of competence in his ability to read and understand academic papers.

To be frank, I'd argue his white paper is plagiarism. I'm surprised Standford hasn't penalized him for this yet. It really looks bad on them. His white paper is laughable. Sample size is from a biased source, and no definitive way to verify if participants tests even match Sharur's target. It has already been thrown out, what more does one need to see at this point.

His latest spiel is how he claims 11mm and 8mm coaxial drivers effectively yields you 19 mm of driver space. His understanding of basic audio physics is comical, and I can understand why he isn't fruitful in his pursuits. He doesn't add, just regurgitates what is already known and claims to have invented it, while acting arrogant and malicious to others. His hard-on for Crinacle and his rejection tilted Sharur's ego, which makes me question whether Sharur suffers from Narcissistic Personality Disorder. If you've ever partook in his community, the fest pool of filth that goes on in his channel is outright disturbing, if not criminal.

So, again, what is constructive here?

 

[Eddie_Vanjovi]

Well, I think he's on to something. On my second try, I noticed the sliders don't exactly do what you intuitively would expect them to do, so I stopped following my preconceived notion of what my IEMs EQ curve and sound should be (what I usually EQ to), and started actually listening to the songs, trying to find what would strike me as pleasant sounding.

My IEM of choice is the Letshuoer Mystic 8, which is definitely overpriced in today's market, but whose quality and comfort stops me from selling it short:


View attachment 487559

My preferred EQ Curve for it, which I came to after extensive testing and biased adjustments using squig.link (I seldom use AutoEQ, and I know my ear modes by heart),looks like this on Peace
View attachment 487574

And here is PEQdB's suggested 15 band curve:


View attachment 487564


Below, what would both curves would be following squig.link's predictions (red being PEQdb's/Sharur's)

View attachment 487571

Case and point. The EQ doesn't even match PeqDB. The whole "study" is a farse. This is why we peer review.

 

[Curupira]

Case and point. The EQ doesn't even match PeqDB. The whole "study" is a farse. This is why we peer review.

By PeqDb you mean the target curve or the Dynamic EQ? Because it kinda of hinds in the same direction of the Dynamic EQ, although, true, it is just one biased data point.

Regarding his target curve, yeah, I didnt like it, it´s too bright for me.

 

[markanini]

The affiliation is real, but it’s not Stanford University proper like some people might think. Saurav Chala is a Master’s student at CCRMA - Center for Computer Research in Music and Acoustics, under the School of Humanities & Sciences. CCRMA is a legit Stanford center, but not a core academic department like EE or Psych. The talk was a regular CCRMA Wednesday colloquium. It’s not an endorsement by the broader university, and it’s not a PhD in hearing science. His whitepaper, the one tied to PEQdB, isn’t peer-reviewed, part of any Stanford-affiliated research, or formally backed by the institution. It's just a company doc from a student.
He has no academic publications on record, normal for most Master’s students, but it does mean there’s no traditional research credential behind the claims. PEQdB Inc. is registered in Delaware with a Stanford, CA mailing address - student housing, which blurs lines in a way that feels like marketing. Stanford peers probably aren't clued into his online rep, past forum bans, uncredited use of other peoples measurements, childish attacks and drama targeting Amir, oratory1990 and Sean Olive. So, no sign the school condones any of this beyond letting a student present.

 

[Alexander Gödde]

His latest spiel is how he claims 11mm and 8mm coaxial drivers effectively yields you 19 mm of driver space. His understanding of basic audio physics is comical, [...]

I saw that in one of his videos. "Comical" is the right word, and it's not even about audio physics but simple geometry in this case. Plus he is the textbook definition of a toxic personality, liberally mixing his criticisms of others with personal attacks.

 

[dasherzx]

I used peqdb to tune the ether cx to my liking. the preset eq on the site is too bright for all the headphones I own. i think it wont be accurate too since the same models are not guaranteed to measure the same. still, great tool and already more value to the community than 90% of audio channels out there. theres alot of ragebaiting but in the end all it mattered was me getting sound i want.