The Violin Doctor | Hacker News Reader

The Violin Doctor

chicagomag.com

49 pointsSirLJ33 comments

The Fine Arts Building in Chicago's Loop is beautiful. It houses a theater, a ballet studio, luthiers, music teachers, a bookstore, a sheet music store, and various others. Rents are pretty reasonable. The elevators are all manually operated.

Absolute experience for me was just taking the elevator to the sheet music store and then walking down the stairs and reading the signs of the various renters on each floor. Music classes, instrument shops, so much culture in a single building.

Sounds like paradise. And sheet music stores are increasingly rare.

> Scientists have used 3D lasers and CT scans to measure the precise geometry of a Stradivarius, but so far no one has been able to replicate its sound.

This is demonstratable false. From https://en.wikipedia.org/wiki/Stradivarius#Comparisons_in_so...

> The Stradivarius instruments are famous for the quality of sound they produce. However, the many blind experiments from 1817[4][5] to as recent as 2014[2][6][1] have never found any difference in sound between Stradivari's violins and high-quality violins in comparable style of other makers and periods, nor has acoustic analysis.

A Stradivarius is a high-quality instrument, but there is nothing unique about it sounds compared to other high-quality violins. The value comes primarily from its artistic and historical nature rather its sound quality. There are modern instruments which sound better even if they are not as valuable.

I'm a double bassist, and other family members are string players. I think there are problems beyond replicating famous instruments, especially for those of us who are likely to play "lesser" instruments throughout our lives / careers. One is that instruments are actually designed to sound different from one another, and it's up to the player to decide what they want.

When I was shopping for my current bass, I was able to play instruments alongside one another that had noticeably different characteristics, along with different design details such as dimensions, and price tags. There is no "best" for makers to aim for, and among the choices that we can actually afford, there are real tradeoffs that have to be made. A wonderful sounding instrument might have maintenance issues that make it an expensive headache to own. I steered away from those basses regardless of sound.

And there's no practical way for an individual musician to conduct a blind test or compare measured characteristics, though I would absolutely love it if there were. Most players would like to cut through the bullshit and woo.

And don't get me started about bows and their maintenance.

Oddly enough, Stradivarius could not have known how his fiddles would sound today, because most of them have been modernized.

> One is that instruments are actually designed to sound different from one another, and it's up to the player to decide what they want.

So true

> And there's no practical way for an individual musician to conduct a blind test

Also true. You can try to approximate. When my kid was younger, I was with them at a violin lesson and the student before had three different instruments they were auditioning for their first 4/4-size violin. So being a full-size violin, this one was to be their instrument for the next few years. The shop had let them take 3 instruments in their price range on loan for a week to audition. The teacher had the student turn their back and labeled the instruments A, B, and C, and proceeded to let the student listen to how each instrument sounded as the teacher played with a couple of different bows. We all kept notes on our favorite, but didn't tell the student yet. The student's mom took all our ballots to consult later after the student was closing in on their own choice.

Over the week the student picked an instrument and had it at their next lesson. The first thing that the student played as a warm up was a 3-octave G-major scale. I said: "Ah, you picked the Dunov, that was my favorite, too." The mom looked at my like I had grown a second head, but to me the instrument's sound was very distinct from the other two. Instruments do vary a lot. They all have their own personality and it is fun to find it -- two things I miss most about kid having grown up and move out are a) listening to practice every day, and b) violin shopping.

Yes, all Strads are Frankensteins in some way, and for some a ship of Theseus problem emerges.

Which makes the entire question of which is better an odd one and quite frankly kind of uninteresting.

> Oddly enough, Stradivarius could not have known how his fiddles would sound today, because most of them have been modernized.

If they weren't, would the tone change as the materials aged?

Yes, according to conventional wisdom. Also, that change is thought to be accelerated by playing. It's thought that a brand new instrument needs to be "played in" for at least a few months, in order to discover its long term potential.

Note that I'm being deliberately non-committal here, because some of this stuff could be bunk or overrated. I don't think my new bass has changed in 10 years. But my ears and my playing have changed.

On the other hand, a modern violin is considerably louder than an un-modified Strad, of which there are only a few, and they tend to be museum pieces. If you want the sound of a Baroque fiddle, there are cheaper ways to lay your hands on one. ;-)

Bows, and E strings on violins. Black arts.

I don't play violin, but I play guitar and have built guitars. There has always been a lot of magical thinking about vintage guitars (and the very specific species of rare woods that they are made from) that just doesn't seem right to me in my experience. It doesn't surprise me that the same thing happens in the violin community.

That said, I'm glad that these historical instruments are so carefully preserved. Violins are very delicate, and to compensate for that, they designed to be almost infinitely repairable. But that only happens if they are perceived to be significant enough to justify the work.

The good part is even cheap violins are repairable and thus there is plenty for amateurs to praction on before trying something that matters.

I just finished fitting a new bridge on one of these. Lets just say I need more practice.

Good setup and playability are huge. I got my kid a "beach and camping fiddle" -- this is not the instrument you use at recitals, this is the instrument you take for playing tunes around the campfire and isn't going to be materially damaged by a beer spill. Good strings, lower the nut for playability, carve the bridge and set the sound post properly, and you end up with a decent fiddle. Doesn't sound like the "good" instrument, but is still fun to play and sounds good enough for some fun.

Carbon bows are getting darn good, too.

As a violinist I disagree with your conclusion. There are a lot of valid criticisms about those studies, including some from musicians who participated. There's also a notable conflict of interest in that the PI is a violin maker and has a vested interest in proving the quality of modern instruments. I summarized the research and criticisms in a Reddit comment[1] a while back.

There are so many subjective components to sound preferences and value it's simply impossible to make claims like "nothing unique about it sounds compared to other high-quality violins", "value comes primarily from its artistic and historical nature rather its sound quality", or "There are modern instruments which sound better"

[1] https://www.reddit.com/r/askscience/comments/ug0an2/why_cant...

One not-exactly-related thought is that with violins, both the room and the bow have a very large effect on the sound; it certain cases, it might be quite difficult to tell the sound of violins apart because the room itself is having an outsize effect on the sound. This doesn't by any means reject the findings of the study, was just a thought. Source: violinist for 13 years, currently attending a top US music school (for composition, not violin, but I still play in ensembles and take lessons).

Happy to try to answer any violin questions!

> Happy to try to answer any violin questions!

As a math grad student at Indiana University, liked music, had recordings from Vivaldi through Rachmaninoff, had a dorm room next to the music school, got started on violin, and eventually make it through a few pages of Bach from the E major Preludio and the Chaconne.

Now I'd like to get back to violin, and wonder, what strings should I get? There at IU, I was told that Piastro Eudoxa strings were "good" so always used those.

What are some more recommendations?

I personally use Dominant strings for A, D, and G, with Pirastro Gold for the E string. Besides the E string, I haven't experimented with much else. If you're just getting back into things, it shouldn't matter too much; Pirastro Eudoxa also seem good. You likely already know this, but if your violin has fine tuners, you probably want to go with ball end strings instead of loop strings, and make sure to get the right size (4/4 if you're on a normal full-size violin).

Thanks. I'll keep that and use it!

My violin is a decently good one,grown up size! And I have a fine tuner only on the E string.

Is the experience of playing a very expensive violin more enjoyable? Is it easier to express the music on such an instrument? I ask because, for me, a pianist, an older Steinway B that has been meticulously maintained is so much … nicer, even easier, to get the sound I want out of the instrument, when compared to almost any other piano (with some exceptions, such as Grotrian, some Mason & Hamlins, Fandrich & Sons Custom and a few others).

As someone who plays violin and piano, it's a similar experience. The more expensive violins will certainly sound better (at the minimum, they are usually more resonant), and it's much easier to evoke the emotion you'd like. The diminishing returns hit hard though, above a certain price you're paying for the history and brand.

The steinway's are wonderful instruments aren't they! I've been dreaming about it and recently had the opportunity to buy a personal NY model B and I don't regret it. The joy it brings you when complex sounds are so easy to get out the instrument, it's priceless.

It's a law of diminishing returns type of thing. I haven't played on any super super expensive violins, but there is definitely a big difference between ~$100 violins and ~$1000 violins. Generally, the more expensive the violin, the more resonant (as opposed to squeaky) it is.

Maybe I should have phrased it as “exalted” rather than “expensive.“

I play a new Bosendorfer and it’s a crazy good experience. Nothing like a Steinway though. I just don’t like Steinways but obvs not dissing them.

Violins feel markedly better at about the $3K, $5K, and $15K levels, but I’m so bad at violin that my $1K antique is perfectly fine. But bows, strings (and separately the E string) also play huge roles. Much bigger than say guitar strings for that instrument.

>Scientists have used 3D lasers and CT scans to measure the precise geometry of a Stradivarius, but so far no one has been able to replicate its sound.

To my understanding, most modern violins are patterned after the Stradivarius. Modern violins sound the way they so because of all the violins of the early Baroque, only Stradivarius’s violins (and to a smaller extent Guaneri’s) survived to the Classical and Romantic, where (with some modifications) became the Platonic pattern all luthiers followed for violin shape, size, and timbre.

So in a way, all modern violins replicate Strads because that is the only violin sound in existence.

For the Bell violin restoration I'd also be interested whether anyone would be able to detect a difference before and after. It shouldn't be too hard to teach a robot to bow a few notes and record whether it changed after he was done fixing it.

Would only work well for open strings, though, once the player puts the finger on the board all sorts of instantaneous feedback happens...

You would need to teach the robot how to shift balance from where the index finger is placed to where the pinky is placed on the bow as the bow travels up and down the strings so that a consistent pressure is placed that produces the silvery, warm tone violins are best known for.

Violinist here, bowing also includes a feedback loop so I don't think a linear mechanical action could indicate how it sounds with a human player. But my signals background says can't we just hit with a hammer and measure the impulse response?

yeah, impulse response would definitely be a good starting point.

> It shouldn't be too hard to teach a robot to bow a few notes

Uh no. Producing a good tone on violin takes months to start getting right. Holding the bow right even for open strings is nontrivial.

Not saying a good robot violinist is impossible. It will happen in our lifetimes, even though I’m old. And I do like your concept of an objective player. But this is not in the “shouldn’t be too hard” category.

IMHO, for music, e.g., violin music, better than a robot playing a real violin is just a computer with some software, apparently still needing development, making the sounds of a violin, piano, orchestra, etc. I think of this future much like typewriters getting replaced by computers and word processing. But in both cases there are humans involved in the performance.

BUT: There is a lot of culture surrounding music from its performance, and tough to get any of that from a computer performance. So, grand concert halls, orchestras, conductors, violinists, and famous, old violins have not much place in that future.

But, to some extent, some of the music, even if performed by a computer with software with a human using the software as a tool does have a future.

Theres quite a few parameters to evoking the sounds and emotions you want from an instrument. We've been able to do it on computers for years, thats what vst's are after all (and modern modeled piano vst's sound incredible). In the end, playing a midi control is still easier than inputting all of those parameters manually though.

> But, to some extent, some of the music, even if performed by a computer with software with a human using the software as a tool does have a future.

Pretty much every modern song is made this way, isn't it? Apart from "live" recordings I'd wager every piano pop song in the last decade was recorded on a VST with a midi controller.

I did a grade F--- job of explaining my thinking.

You mentioned "VST". Never heard of it. I just looked it up and found

Virtual Studio Technology

https://en.wikipedia.org/wiki/Virtual_Studio_Technology

So, there is a "synthesizer"; sounds nothing like anything I have in mind.

To try again, I really like music and have for decades. The music I like is nearly all classical from between Vivaldi and Rachmaninoff. I liked music so much, if only to learn a little more, while I was a math grad student at Indiana University (with a terrific music school) I started on violin and made some progress on, e.g., the Bach Chaconne, and that is significant work.

The music I like is art with the short definition

"Communication, interpretation of human experience, emotion".

Since no machine can have a "human experience", essentially no machine ever can create such art.

Well, I'd like to pick up my violin again and get back to music as art, but (1) getting good at violin, good enough to "communicate human experience" is not easy, and (2) I'd like also to move past just solo violin to, in two words, a symphony orchestra.

So, I'd like to write, create the music and perform it. Hiring a symphony orchestra would not be cheap.

Every performance of music that is on a DVD has been reduced to just bits, and in terms of current computing, not very many bits. So, necessarily, in principle, a computer could generate such bits, not create the music but just generate the bits.

So I want to use a computer to substitute for part of what an orchestra does. But an orchestra has good musicians that do real art, and, again, a computer can't do that, i.e., we can't hope for software that we can ask to create and give a good performance of something as good as the best of, say, Dvorak.

Well, a violin by itself can't create art, either. All the art has to come from a human using a violin to make the sounds. So, I'm trying to be clear here, I want a computer to substitute for a violin, etc.

For MIDI, I have no interest in that. From all I have seen about MIDI, it is hopeless for the level of art common in orchestras and that I have in mind.

Maybe I've been more clear here: Again, once again, I want to use a computer to perform the music I have in mind.

I don't think the software for what I want exists yet.

And for whatever has been done with pop music, I have no interest in pop music. To me, pop music is nothing like the art I have in mind or saw in the best from Vivaldi through Rachmaninoff. So, in particular, I doubt that I would be interested in anything about computers that have had a role in pop music.

Or, for violinists, I want an expressive performance of the last bars of the D major section of the Bach Chaconne. The performance I have in mind requires careful construction of each note (volume and the nature of the sound and how those change during the extent of the note) and transition (e.g., legato) from one note to the next. A violinist might work for hours to months on just the legato on just a few bars of the Chaconne -- for what I have in mind would take comparable effort. And then move on to more, already in sheet music or that I write.

E.g., for some art, an example of expressiveness, I like the performance at the beginning of the second movement (Adagio) of the Dvorak Cello Concerto (opus 104) as played by Rostropovich, the Berlin Philharmonic, and von Karajan. I didn't see this on YouTube.

For something expressive on YouTube, I like (parts of) the Nicola Benedetti performance of the Bruch Scottish Fantasy at

https://www.youtube.com/watch?v=YpX8DoS2hr8

In addition, the "Stradvarius" sound has been traced to the density of the wood due to the prolonged droughts that produced it. If we use similarly dense wood, modern violins have similar "quality".

Furthermore, almost all of the Stradivarius violins have had some level of modifications or repairs to bring them up to a more "modern" standard. Is that still a "Strad"?

In addition, wood is known to degrade over time. No matter how good the Stradivarius violins were, the longer they exist the less good they become.

Reply on news.ycombinator.com