• Content Count

  • Joined

  • Last visited

About notsodeepblue

  • Rank

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

  1. There's a fantastic picture on Guy Harrison's website here that gives some insight into the illustrated scrapers.
  2. They can also be seen here, as well:
  3. I agree completely, but think there's a fair bit of value in an interactive 3D model - being able to interactively slice in x, y and z is something I find extremely useful. Have you looked at converting your Matlab-generated model to a web-friendly format, and getting it into the browser (and possibly sharing online)?
  4. ...I thought they only got used as makeshift goalposts when it was too cold to take a coat off, but I can see that would work too.
  5. I think this is a really important thing to look into before you decide what you want to do - attached a picture of the packaging of (vintage?) P&H DIY hair I have lying around: "With Horse Hair" is not exactly confidence-inspiring, and the included hank looks truly awful. While this pack is of an unknown vintage, it does not appear that P&H were putting much effort into the product beyond the marketing. Iit almost seems that you are asking the question: "What matters more - the stick, or the freshness of the hair on it?". If so, you should probably start to factor in the quality of the hair and rehair as well - the two options you outline are as different to one another as your choice of bow-sticks.
  6. Perhaps it's just me, but does the top not look to be too far off-quarter for dendro to be used confidently in this case?
  7. This has the feel of an undergrad or masters level research project - is that the case? If so, you might want to talk to your supervisor before doing anything further with a dataset produced as part of such a project (they may have the idea of having other students build on your work, and would potentially take issue with the raw data finding its way into the wider world if it limited those options...or maybe not, but best to ask anyway if that's where your data comes from). Any chance of posting a couple of training images at the resolution they were fed into your nn (ideally from opposite ends of the price spectrum)? It would potentially be a good ground-truthing exercise to get a sense of whether your network is being trained on images that even contain features that would be considered germane to the violin identification process before deciding on whether to invest more time in your outputs (i.e. if not, you will struggle to convince a large chunk of the potential market for your tool/data that it can actually prove useful).
  8. I can see that at this level, a shop would have nothing to offer you in reality. I wouldn't be too quick to dismiss a local violin shop as a good (occasional) source of such instruments: I visited my local violin shop for some strings a while back, and after introducing myself and explaining my interest I also ended buying one of the numerous "project" violins the owner had, which he was confident he would never get around to working on and wasn't going to miss. I would also argue that if you go down this route, you are also getting the benefit of a professional having seen the instrument and deciding it was repairable - the ebay route leaves you with only with a handful of pictures, and your opinion. And as others suggest, this kind of relationship can prove invaluable in the longer term: While it's not impossible that you might get this with an ebay seller, it's significantly more likely you will with a person standing in front of you who shares your interests.
  9. Just subjected the rest of the original spindle to the sniff-test, and no char (or anything else). I don't think I can recall anything of the like when I received it either - thanks for the suggestion though, a readily believable cause of such variation. At which point, I think I will just accept that this is meant to remain a little mystery and leave it be. I will keep my windowsill / dark room control samples where they are for a few weeks, and may post a final picture if it produces an impressive contrast just for posterity. Thanks all for the ideas and thoughts - I've picked up all sorts of interesting links and suggestions as a result.
  10. Thanks for the recommendation - I'll give those a go.
  11. Thanks. The listed density definitely seems to be in the right ballpark. Does Mpingo undergo the type of rapid colour change this wood seems to be capable of? That's something I've only ever associated with pretty looking veneer, and so didn't really consider it. I will go and browse some images - thanks. As a follow-up, after 72 hours sitting on a south-facing windowsill - exposed to occasional, watery English sunshine - my mystery wood is now looking like a rugby jersey: I haven't been able to get hold of anything potent to test it with yet, but did confirm that potassium permanganate had absolutely no effect on a freshly scraped section.
  12. What a fantastic resource - thanks very much for the link. I can't imagine that the spindle I bought could possibly have been that old, but it was in an online wood-dealers "specials" category so perhaps not something they knew too much of the history of either. I have no ammonia or acid to hand, but will do my best to track some down - I will put a drop of whatever I find on a freshly-sawn and scraped face, documenting the before and after whenever I am able to find some (might take a while, but I will make a point of doing this whenever I am able). For the interested, I attach a picture ~24 hours on from cutting (sorry for the slightly washed out colours on the right hand side of the image). The top blank is the colour as-received (i.e. outer surface colour of the original spindle), the middle is with me chasing direct sunlight for a day (6-8 hours total drect sunlight, the rest either indirectsunlight or darkness), and the bottom section has been sitting in a darkened corner of a dark room for the same amount of time: I will keep these two pieces where they currently are for a couple of weeks at least, and update the colours if and when they change significantly. thanks all,
  13. Thanks - that was my first instinct when I saw how substantially different the interior was to the exterior. This timber is so dense (this piece is ~1,200kg/m^3 as-received) that that I thought it might have been a cheaper / less desirable material treated to deceive, somewhere along the supply line where volume or weight was involved. I have taped-up a section of freshly exposed wood and left it in direct sunlight to se how much and how quickly this wood tans - if there is an interesting effect, I will post a picture of the results. thanks again,
  14. That is such a stunning bow - congratulations!
  15. too! It is indeed both, and attached a shot of a rough cut blank that I think shows this better, as well as how the colour bleeds into the surface a few mm and in places more deeply (the cut-surface is straight off the coping saw, no planing or scraping at all): The dark outer surface is the same as when I received it a couple of years ago, and doesn't seem to have changed much when kept in indirect sunlight, along with other timber. It was a 50x50mm spindle that was uniform on all 4 faces of the blank, hence when I saw the colour of the dust while resawing it I became ...confused. I am going to be putting a small sample under uv light later this weekend to see if / how quickly this transformation occurs, but it is a pretty extraordinary difference and hence I was struggling to identify from photographs what it might be.