Again culling from emails sent in my absence, there seems to be significant cost disparity when getting price quotes for having patterns digitized versus the final bill so I will try to clear up some of the confusion. The first confusion relates to per piece pricing, the second can be related to the CAD program (or scanner) your service provider is using and the third amounts to preparedness. If you don’t prepare well, it will cost more.
Per piece pricing:
Contrary to many assertions, digitizing should not be time consuming and expensive -but everything is relative. If you wanted a simple 5 piece pattern digitized, it should only cost $15 to $30. However, since this is a small job, you may have to pay a minimum shop rate. For better or worse, this is akin to calling a plumber to repair a faucet. It may only take 15 minutes but a plumber has to charge a one hour minimum if only due to opportunity costs. Meaning, in accepting your job, the plumber has to turn down other jobs that would likely pay more. In sum, if you only had the one small job, you may be on the hook for a minimum charge ($50-$100). It would be unkind and unfair to suggest a service provider is charging you $10-$20 per piece when what you’re really paying for is a minimum job fee.
Solution #1: It would be better to send more patterns (if you have them) to use up the rest of that hour to lower your per piece cost. If your job ran over the hour, you’d be paying closer to a standard rate of $2 to $5 apiece. For more on the going rate, see the link below about sending patterns off for digitizing.
Software related costs:
Some software programs used by pattern makers and graders are not very efficient and take significantly longer to digitize pieces. The program that stands out in this regard is Gerber; it is much more laborious to digitize with (comparatively) and also, the Gerber software pattern file management system is very inefficient (this is no insult to Gerber users, you do what you have to do). Other programs such as StyleCAD (which I use) and Optitex are better. That said, the advantage of having your patterns in Gerber is that due to market dominance, there isn’t a plotter or cutter on the market that can’t process these files and probably natively. Not being able to use Gerber formatted dxf (trust me, not the same thing as generic dxf) would be like manufacturing PCs for the average user (note: average user) that couldn’t run Windows.
There is a new technology called N-scan (and probably others) that appear to be more efficient than any CAD program but after having worked with a few scanned files, must report this is relative. Digitizing a pattern amounts to feeding a pattern piece through a slot, its dimensions are read and captured quickly. However, these files require a great deal of editing after the fact. The reason is that the scanners insert a very large number of points on each line. Managing points is resource intensive and can complicate grading later on so many of these points need to be deleted through manual editing. This is also true of Gerber; it requires a comparatively large number of points.
Solution #2: All other things remaining equal, select a provider who uses more efficient software but who has the capacity to export Gerber compliant output. A CAD company can’t get any traction if they can’t export Gerber cut files. If a program can’t write to Gerber, it is not an industry CAD program no matter how much its creators claim to the contrary. There are links on this at close.
The costs of preparedness:
The cut to the chase summary is that each piece you send must have an information block that contains all pertinent information such as grain line, piece names (which will be changed in digital format), number to cut, size, material (shell/self, lining, fusible etc) in addition to a pattern card. Also, each style must -MUST- have a style number. You also need to include the seam allowance for each seam and last but not least, you can’t use sharpies to trace the pattern pieces because the lines are too thick and dark to read through the puck of the digitizing stylus and will need to be retraced or worse, guessed at.
You also need to be familiar with conventions used in production pattern making (see pgs 176-180 of my book) such as color coding and to include knowing that any drills are typically sewn 1/2″ beyond the point (if you’re not certain, trace out the finished dimensions of the dart and indicate these are finished dimensions). If you don’t and dart ends come out wrong, we are not at fault. We don’t know you don’t know since nearly everyone else does. By the way, those little cut outs along the edges are very important too. We call those notches. They are not a superfluous decoration; don’t skip them.
If there are problems with the patterns you send in that not all the information needed is available, it will involve a telephone conversation during which you’ll have to retrieve your pattern, your sample, a ruler or tape and what not to answer the questions which increases the amount of time spent on your project. A service provider can’t stay in business only charging $3 to digitize a piece that requires a 15 minute conversation. Multiply that by an hour and the service would only earn $12 for four pieces which is well below their hourly.
Solution #3: Be prepared. Lack of preparedness is the biggest cause of cost over run in pattern digitizing.
Related:
Sending patterns for digitizing (skip this to your peril)
Sending patterns for correction
Why pattern makers don’t want to grade patterns
How to know if you need digital or paper patterns
CAD software compatibility in marker making
Checking a pattern pt.1
Checking a pattern pt.2
Hi, I am at the company that makes N-Scan, you mentioned it. Thanks, it is a nice report, since we focus on digitizing (the title of this blob). I am wondering if you have used our system? I ask you since you can set up the system to automatic detect the curves/lines detected with more or less points (with the understanding that with fewer points the less accurate is the curve, just like with a human digitizing in a table). So with the parameter adjustments done once, before the NScan starts to be used, all customers adjust to a quantity of points that they are comfortable (and the NScan will produce the best accurate description of the lines with such an approximation, using an artificial intelligence method to select the points better than a human digitizing does). This is all automatic. So the customers get as few points as they would get by humans using a digitizing table , but with a typically better curve quality (smoothness) since the points are chosen by a good artificial intelligence program (just ask any customer).
I gave you my e-mail above if you have any questions. Thanks again to refer to us
best regards
-Davi
Hi Davi
No I haven’t used your system. Digitizing isn’t a big enough side of my business to buy it altho I suppose that could change. I reviewed what I could of it here after seeing it at the Spesa 2007 show. The lady who demonstrated the product for us didn’t understand why I was asking all the questions I was even though I didn’t intend to buy it so she wasn’t very helpful. I don’t think she understood what blogging is or the reputation the site has in the community. After I published my post (giving your product a good if not excellent rating), she emailed me a less than hospitable email so I haven’t looked at it since.
Anyway, it is good to know the system is configurable with respect to points which could reduce editing downstream. One style I received that was scanned with the system had so many points that it crashed my program. I will say it was a very complex pattern so I ended up separating the pieces into different files so I could edit each piece individually and then put it back into one file once file size was reduced. This was just the one style, I’ve had a total of maybe 20 patterns scanned with N-scan so other people with more experience could probably offer better insight.
I would think the system configuration could represent conflicts in some ways. By that I mean that a service provider would be motivated to keep as many points as possible in the interests of accuracy and the client might want that too although it would cost them more down the road with editing. Probably the best balance is achieved with a provider who has a lot of experience with CAD programs in that they’d understand the minimum points necessary for accuracy. I think we could all agree that no line, no matter how complex, needs a point every 32nd of an inch!
Hi Kathleen,
thanks for the comprehensive response. Let us know next time you need a pattern digitzed, we should be able to help you find near you someone using the NScan. I am sorry the bad experience you described, I am ready to fix it :-) . Also, I understand the discussion you present about service provider and clients. Let me point out something that only the NScan does: the NScan can save files in various different formats (virtually every CAD format) so one can have simutaenously multiple file formats for each pattern. What the NScan also does, is to select points according to the file formats. Say someone uses Gerber (as you discussed in your blog) or Lectra. Each CAD completes curves with different methods. The NScan will save curves with different artificial Intelligence methods, so that the curves can be recovered accurately in the respective formats.
This is indicates how careful the NScan is with selecting points with artificial intelligence. This abilities can help Service Bureaus to save files accurately, with fewer points, according to the client format requirement.
Thanks for giving us a chance to discuss with you about our system, and please, be in touch with us to have patterns digitized by the NScan.
Hi,
I’m looking for digitizing and layer marking of my manual patters. We are working for Gerber old version software. Can you help me in this regard. We can hire someone for part time with us who can work on that software or it can be done by third party on per pattern price.
Please advice if you can help in this regard.