acsearch rate hike

Discussion in 'Ancient Coins' started by lrbguy, Jan 10, 2019.

  1. lrbguy

    lrbguy Well-Known Member

    I just got the email letting us know that their (acsearch) new search engine has been beta tested and is ready for use. Mostly the same bells and whistles, but now they will charge .70 EUR per search. Paid subscribers will get 20 freebies before the search fee is charged. Peons (like me) will pay as of now. After currency conversion Yanks will pay 80cents per search. Our Canadian friends will pay just over a dollar per search (not per session).

    "UNCLE" he cried.
     
    Alegandron likes this.
  2. Avatar

    Guest User Guest



    to hide this ad.
  3. TIF

    TIF Always learning.

    I think you've misunderstood. This new feature is good news (well, it would be better if were free but still...)

    In addition to using the database as we've always done (free searches for everyone, but only subscribers can see prices realized), now you can do an image search! That's what they're charging for-- an image search. I wonder how well it works? I'll try it tonight and see.

    It's far cheaper than Ex-Numis although I don't know how their image recognition software and databases compare.

    Here's the announcement from their website:

    Image search released

    10.01.2019, 22:12

    We are pleased to inform you that the long-awaited acsearch image search has successfully completed the beta phase and is now available to all users. With this great new feature you can quickly and easily find entries of the same coin or die-matching coins in our archive. Areas of application include pedigree research, die studies, coin attributions, etc. The search works in real time and results are displayed within a few seconds. For an optimal search performance, please read the usage instructions.

    Due to the complex and expensive infrastructure, the acsearch image search is subject to a fee (starting from 0.70 € per search).

    All premium users receive 20 free search credits as a thank you for their support.
     
    zumbly, Ryro and Justin Lee like this.
  4. TIF

    TIF Always learning.

    Here are the instructions. Looks like I'll have to change my images backgrounds to white but that's not a big deal :).

    ...

    Usage instructions

    The image search engine offers a quick and easy option to search for entries of the same coin (f.e. for pedigree research) or for coins struck from the same dies (f.e. for die studies or coin identifications).

    Please note that for coins struck from hand-crafted dies (ancient and medieval), identification searches are usually of limited use as only few die-matching entries will be in the database. For coins made from machine-made dies (modern times), on the other hand, the use of the pedigree research is rather limited due to the high number of coins struck from the same (or very similar) dies.

    Since acsearch contains professional coin photographs for the most part, the best search results are obtained by using a professional coin image for your search query. Please note that:

    1. The image should contain both sides of the coin.
    2. The background of the image should be white (no patterns and no textures, e.g. by cropping in Photoshop, Gimp, etc.).
    3. The image should be cropped and not contain a frame.
    4. The shadows on the image should be as natural and unobstrusive as possible.
    Snapshots from a smartphone are often too dark and blurred and, therefore, not suitable for the image search engine. In general, the better the query image, the better the results.

    If the image search engine returns many results, they can be further narrowed down using the text search function. For coins from dies that differ just marginally from others (e.g. same image, different date), the date can, for example, be used as an additional search criteria.
     
    zumbly, Ryro and Alegandron like this.
  5. Orfew

    Orfew Draco dormiens nunquam titillandus

    I agree with @TIF this is very good news. I have 20 credits as a subscriber and I will certainly use a few tonight to see what results I get. Hopefully, this will be a great tool for finding lost provenance.
     
    Ryro likes this.
  6. TIF

    TIF Always learning.

    I ran an image and it did find a prior sale that I was unaware of and it found the listing from when I bought it. However, I could have just done a regular search and found the coin.

    The coin:

    [​IMG]
    MOESIA INFERIOR, Nikopolis ad Istrum. Elagabalus
    Æ 26 mm, 12.5 gm
    CE 218/9, Novius Rufus, consular legate
    Obv: ...ΠΗ ANTΩNEINOC΄ laureate, draped and cuirassed bust of Elagabalus right
    Rev: VΠ NOBIOV POVΦOV NIKO[ΛOΛITΩN ΠPOC IC TP]; Athena standing left before serpent coiled around olive tree to left; shield on ground behind her
    Ref: Hristova & Jekov 8.26.4.7 (same as last illus. Specimen); Varbanov 3906 (same dies as illus.); Moushmov 1384

    The prior sale 2007 (listing; it was unsold on an overly optimistic estimate) which was not listed when I bought the coin, although I may have seen it and not noted it when researching comps:

    https://www.acsearch.info/search.html?id=391231

    The listing from when I bought it:

    https://www.acsearch.info/search.html?id=3669467


    The image search has the potential to be valuable but will depend on ACsearch adding thousands of old catalog images and data. Even then, if a coin is not a super common type, you can still just search the regular way-- it just takes more effort. I can see how it might be easier to search for prior sales (and possibly die matches?) for coins that were issued in huge numbers-- it would save you from manually looking at every image.

    Perhaps it could be of some use for identifying unknown coins. That might be interesting to test.
     
  7. Orfew

    Orfew Draco dormiens nunquam titillandus

    I just used 2 credits to test the image search. In the first case it found no match, but in the second trial it found 2 matches (the coin being sold in 2 different auctions). I already knew this information from a regular search but it was neat to see it work.
     
    Justin Lee likes this.
  8. dadams

    dadams Well-Known Member

    Has much is a premium subscription? I've used acsearch but only as an unpaid visitor. The image search seems like a useful function.
     
  9. TIF

    TIF Always learning.

    Hmm, that information is surprisingly hard to find on their website. Here's a screen shot of the rates:

    Screen Shot 2019-01-10 at 9.41.13 PM.png

    I bit the bullet and subscribed soon after they began charging. It is absolutely worth it to me. While you can still search without a subscription, finding the hammer prices is time consuming if you don't have a paid subscription. You have to look up the results off site, either directly with the auction house or with Numisbids or Sixbids. Having the hammer prices on ACsearch saves a lot of time. The new image search service is intriguing too and having 20 paid searches with my subscription is nice.
     
    Orfew and dadams like this.
  10. dougsmit

    dougsmit Member

    Ex-Numis had two useful features. 1. You could search and find out that they had no information (was it free or cheaply?) and pay for what they found. 2. You could tell them of sales you knew of and not be charged for those. However their fee being charged for each hit found could get expensive when some of us would be happy knowing only the oldest. I have a couple coins I would like to know the results on but never used them due to the price. I really like acsearch and am glad they did not cancel the free part we have had. The value of the image search is not high if it only covers sales already available in the free database. Die matches would be nice but I'd have to see that one proved 'under fire'. I have several die matches that are less than obvious due to centering or wear. I would be amazed if anyone could write a program that could match them from two photos taken by the same photographer on different days. I am often wrong. I hope I am.
     
    benhur767 and Justin Lee like this.
Draft saved Draft deleted

Share This Page