Why don't Chinese lasers use printer drivers?

[Ross Moshinsky]

I like to stay clued in so I've been looking into Chinese laser systems since they seem to become more and more popular, especially since Rodne went to China.

My biggest question is, why don't they just develop a print driver? It seems like it would make everyone's life a bit simpler.

[Rodne Gold]

They get the software from the laser control mainboard mnfgrs so take what they can get or what the card mnfgrs use (costs the laser mnfgrs nada for software) - yes a Corel based print driver would be "better" for existing corel/acad/legacy files but it would require "development" and require amendment when the control card is replaced with a newer/cheaper/easier to work one. The laser mnfgrs in China do not develop their own mainboards..they buy em in. It's way better for consumers as upgrades are actually real cheap , round $400 for a new motherboard and LCD panel (the 2 normally go together) and thus one is not reliant on a specific laser mnfgrs card and can easily use any of the systems and software on the market if one wishes as most can be used in any chassis and with any tube or motion system.
At the end of it all , "better" is relative , using a print drivers properties to design in corel etc is actually a "lazy" way and does not allow the design to be used in some other processes.
For EG , most chinese lasers will NOT raster a line with width , they see it is a "vector" line , but a line with width is problematic for a lot of other machinery , like a CnC mill/router or vinyl cutter etc as they too can't do this or use the Corel drawing , a line with width should actually be a closed filled bounded entity for it to be TRULY portable.
I actually prefer the chinese way , we design in Corel and export and can do more in the software they provide than in corel in many ways , like change cut paths , create lead ins and lead outs and a lot of other stuff besides. My low level staff have less issues using the chinese lasers than my mainstreams. the same Corel drawing can also be used on my many other machines without re designing which we have to do if we try use a mainstream "laserable" file designed for the print driver.

[Ross Moshinsky]

Rodney,

The issue with their current method is that it's limited to Corel or AutoCad when people use Illustrator, Inkscape, and several other programs. At times I've sent a job straight to the laser from a customer provided PDF. There is also a big issue with their method because although they don't have to worry about changes when the board changes, they have to worry about when Corel or AutoCad update. In the end the printer driver is a much more user friendly way which is why it's been used by all major laser manufacturers for years.

As for the Chinese way. We use Gravostyle and it's very much the "Chinese" way and I promise, it's full of flaws. It's especially bad with artwork. If a customer provides layered, complicated artwork in vector form, it will import in TERRIBLY. It will take 30 minutes to get it compatible with the software vs just running it straight from a graphics program. There is also a major issue with color and gradients. Completely takes away the ability to do any of that type of work easily and efficiently. So then you're stuck creating a raster image from a vector image. The Chinese software I'm sure can't change the artboard from 72dpi to 300 or 600dpi so you're ending up with issues there.

I've seen the pros and cons of both. The ability to do it both ways is key. Provide the free software, but also make it easy for people to keep doing it the way they have for years and years. I can't imagine programming a driver would cost more than a couple thousand dollars and they'd make that money back quickly even if they sold it as an add on feature for $100-200.

[Zvi Grinberg]

Rodney,

I would like to contribute to the discussion, but it is important to make it clear to every one in this forum, that I am a Universal Lasers distributor in Israel. Please refer to this as a legit professional opinion, and in any case since the presence of my customers here is very rare, I gain no direct benefit from posting here.

I find it discouraging that there is no real developer of chinese lasers. From Rodney's post I learn that every machine is assembled by a different integrator in different ways. There is no "father" to the development and no wonder that in such an environment it will be hard to find dedication in advancing the interface.

The laser mnfgrs in China do not develop their own mainboards..they buy em in. It's way better for consumers as upgrades are actually real cheap , round $400 for a new motherboard and LCD panel (the 2 normally go together) and thus one is not reliant on a specific laser mnfgrs card and can easily use any of the systems and software on the market if one wishes as most can be used in any chassis and with any tube or motion system.

ULS architecture has a much better (IMHO) approach. Controllers are the same for each series, and the Universal Control Panel (that Print Driver software and job manager) works with all series. That means that each enhamcements automatically reach ALL machines in field, no matter how old they are. This is an architecture that ensures the customer that his/hers machine is being constantly developed and maintains value.

Another fact is that the CPU is not more that 400-500 dollars to replace for most series. Just like the chinese...

I actually prefer the chinese way , we design in Corel and export and can do more in the software they provide than in corel in many ways , like change cut paths , create lead ins and lead outs and a lot of other stuff besides. My low level staff have less issues using the chinese lasers than my mainstreams. the same Corel drawing can also be used on my many other machines without re designing which we have to do if we try use a mainstream "laserable" file designed for the print driver.

Corel files are portable and should remain so. If you have variety of machines this is the best way to reduce overhead and management in your business. And in cases such as Rodney's there could be even more than one "version" of chinese machines - a fact that one has to value in time and money.

Univesal as an example can take a thin line and allows you to consider it as either raster or vector. Giving you the flexibility beyond Corel or any other application. UCP also allows for relocation and duplication of geometry in ready jobs (without having to return to Corel/ACAD).

In the Universal way, you do not send any data to the machine. Everything is stored in your host PC. The machine has no internal memory and is directly operated from the PC.
When you upgrade your PC you actually upgrade the machine. You don't have to deal with old memory chips, CPU's and aged electronics that in time tend to be obsolete and force you to make dramatic changes to the way you work (or dump too old machines).

Again, please don't forget that I am a ULS distributor.
I may be biased but at the same time I have and share good experience that might be valuable for you.

Zvi

[Rodne Gold]

Well , none of my "printer driver" laser files are actually portable to other machines , not even other lasers. They are designed SPECIFICALLY for the driver I am using on my GCC's and do not work with my CnC overhead router , vinyl cutter, digital printer without some modification. It's not much of a mission to adapt a design file to whatever machine you are using it on and then get output. My CnC router works with signlab/profilelab , my cutters work with Corel and a Roland driver, my digital printers also use a Wasatch RIP to output. My small Roland engravers us Dr Engrave and My older ISEL engraving machines work with Autocad and a custom driver I wrote..So we used to messing with Corel or other drawing files to get the right output.
Even on my GCC's , firmware changes often as do drivers , we have to adapt to changes when this happens ...

Not sure why this is such an issue.. You will only get true compatibility and transportability if you use the same type of machine with the same driver. so you have the choice , continue with your original workflow with no issues and buy a high priced machine to do so , or save a bundle but have a slightly more difficult workflow or workarounds to make it the "same" as what you got --- up to you!!!
Just an aside: The suppliers of my 2 chinese machines upgraded their motherboards to newer and better versions , they sent me 3 mother boards and 3 lcd panels for free a month after I bought (2 for machines , 1 for spare) , was a snap to change em , update the software etc. all I had to do was send the old ones back , which cost me $100.
The problem really is that the Chinese machines are pretty much not a plug and play newbie friendly machine , but they are in the "affordable" bracket for newbies , but if you want to enter the engraving/lasering field , you need SOME hands on ability. Kind of a catch 22 situation.

[Ernie Balch]

My new Rabbit laser allows printing directly from both Autocad and Corel. It also has it's own drawing program that exports to the laser. I have been designing in Corel and exporting DXF and raster files to the laser drawing program where I set up the parameters. I will have to look at trying the printer drivers with Corel.

ernie

[Dan Hintz]

I find it discouraging that there is no real developer of chinese lasers. From Rodney's post I learn that every machine is assembled by a different integrator in different ways. There is no "father" to the development and no wonder that in such an environment it will be hard to find dedication in advancing the interface.

Zvi,

This does not surprise me in the least. With copyright laws being practically non-existent (or unenforced) in China, there's little reason to innovate. The moment one person comes up with a good idea, someone is around the block making a copy a week later.

ULS architecture has a much better (IMHO) approach. Controllers are the same for each series, and the Universal Control Panel (that Print Driver software and job manager) works with all series. That means that each enhamcements automatically reach ALL machines in field, no matter how old they are. This is an architecture that ensures the customer that his/hers machine is being constantly developed and maintains value.

This is not completely correct, at least as far as I'm aware (so feel free to correct me on this one). While the majority of the parts are shared between the various machines, older machines are eventually set out to pasture and no longer receive the same updates to the driver as the more current machines. I have no objection to this, just pointing out an inaccuracy.

Another fact is that the CPU is not more that 400-500 dollars to replace for most series. Just like the chinese...

Agreed... though I would say that replacing the motherboard and front panel LCD on a Chinese machine gives you a (practically) entirely new machine. With the Western systems (ULS, Epilog, etc.), there are more PCB. For example, there are three boards in my machine that are simply for motor control. Then you have the motherboard. Then you have a couple of PCBs scattered near the motors themselves and the sensors. then you have the PCBs in the laser cartridge. And so on. Replacing them all can get expensive, especially if you're hunting for a problem.

In the Universal way, you do not send any data to the machine. Everything is stored in your host PC. The machine has no internal memory and is directly operated from the PC.
When you upgrade your PC you actually upgrade the machine. You don't have to deal with old memory chips, CPU's and aged electronics that in time tend to be obsolete and force you to make dramatic changes to the way you work (or dump too old machines).

I may upgrade the PC, but I am in no way upgrading the laser. I know this is the line ULS likes to use when comparing themselves against competitors that use on-board memory (e.g., Epilog), but it's simply marketing and no more. My laser does not any work faster with a top of the line computer than if I was using Windows 2000 on an old Pentium. It either works or it doesn't. Sure, an on-board memory chip might fail on a laser's motherboard, but so could any of the other chips... it's simply a repair that needs to happen, and I can guarantee a SIMM will be cheaper to replace even 10-15 years from now than a failing motherboard in the laser. Again, I have no real complaint about on-board versus off-board memory, just pointing out a flaw in the argument.

Dan (a happy ULS owner)

[Ross Moshinsky]

Well , none of my "printer driver" laser files are actually portable to other machines , not even other lasers. They are designed SPECIFICALLY for the driver I am using on my GCC's and do not work with my CnC overhead router , vinyl cutter, digital printer without some modification. It's not much of a mission to adapt a design file to whatever machine you are using it on and then get output. My CnC router works with signlab/profilelab , my cutters work with Corel and a Roland driver, my digital printers also use a Wasatch RIP to output. My small Roland engravers us Dr Engrave and My older ISEL engraving machines work with Autocad and a custom driver I wrote..So we used to messing with Corel or other drawing files to get the right output.
Even on my GCC's , firmware changes often as do drivers , we have to adapt to changes when this happens ...

Not sure why this is such an issue.. You will only get true compatibility and transportability if you use the same type of machine with the same driver. so you have the choice , continue with your original workflow with no issues and buy a high priced machine to do so , or save a bundle but have a slightly more difficult workflow or workarounds to make it the "same" as what you got --- up to you!!!
Just an aside: The suppliers of my 2 chinese machines upgraded their motherboards to newer and better versions , they sent me 3 mother boards and 3 lcd panels for free a month after I bought (2 for machines , 1 for spare) , was a snap to change em , update the software etc. all I had to do was send the old ones back , which cost me $100.
The problem really is that the Chinese machines are pretty much not a plug and play newbie friendly machine , but they are in the "affordable" bracket for newbies , but if you want to enter the engraving/lasering field , you need SOME hands on ability. Kind of a catch 22 situation.

My real issue is, I wouldn't have to work around anything if the manufacturers spent what would amount to probably $1000 to develop a print driver.

Can I work around and use this Laserworks software? Sure. I just don't understand why we should when developing a driver wouldn't cost much.

[Dan Hintz]

My real issue is, I wouldn't have to work around anything if the manufacturers spent what would amount to probably $1000 to develop a print driver.

To be fair (as a software/hardware developer), I wouldn't bother booting my computer for $1k. Driver development is not a weekend warrior affair.

[Lee DeRaud]

To be fair (as a software/hardware developer), I wouldn't bother booting my computer for $1k. Driver development is not a weekend warrior affair.

I hear where you're coming from. But they've already done the hardware interface/control, which is the hard part of the equation. Getting the existing chunk to talk to the OS gracefully isn't rocket science.

[Kevin Huffman]

My real issue is, I wouldn't have to work around anything if the manufacturers spent what would amount to probably $1000 to develop a print driver.

It's cheaper and easier to make 1 software compatible with a machine by including the driver in the program. Then it is to make a windows printer driver that is compatible with ANY software. Because then you add in all these "home grown" version of auto cad or 1 of the 1,000,000 different free ware programs, it's much tougher to keep up a driver that works with all of them.

[Ross Moshinsky]

To be fair (as a software/hardware developer), I wouldn't bother booting my computer for $1k. Driver development is not a weekend warrior affair.

We aren't talking about US labor. We are talking about Chinese labor. Not to mention, we are talking about companies that should have a couple software people on staff. Assuming they make about $15 an hour, I have a hard time believing it would take them more than 67 hours to complete. I bet Laserworks took 10x as much programming time and they give that away for free.

It's cheaper and easier to make 1 software compatible with a machine by including the driver in the program. Then it is to make a windows printer driver that is compatible with ANY software. Because then you add in all these "home grown" version of auto cad or 1 of the 1,000,000 different free ware programs, it's much tougher to keep up a driver that works with all of them.

Are you suggesting making a program like LaserWorks is less work than making a printer driver? I think if that was the case, Epilog and Universal would have done it 10 years ago. That way they can keep you under their thumb easier. By using Corel, you can easily switch from an Epilog to a Universal without any real issues.

As for compatibility; every other US/EU laser company has dealt with the compatibility issues using printer drivers. The great thing about printer drivers is it doesn't take much to keep up to date. Think about your average printer. You plug it in, install the driver, and you're done. Every once in a while you might update the firmware or driver. If you install a new version of Corel, you don't have to get a new printer driver. It all works great. The only time you have to worry about it is when there is a new OS coming out and that only happens about every 3-5 years with Windows. It takes plenty of work getting LaserWorks updated for the new OS, so I can't see how a printer driver is more work.

In the end, LaserWorks looks like a fine piece of software if you're starting a new business with a laser engraver. You don't even really need Corel. You could probably get away with just Inkscape and LaserWorks. As we all know, operating the laser is the easiest part. The design and execution is where things get far more difficult. To essentially force customers into learning a new piece of software just seems a unnecessary when a "simple" printer driver could resolve almost everything.

[Zvi Grinberg]

This is not completely correct, at least as far as I'm aware (so feel free to correct me on this one). While the majority of the parts are shared between the various machines, older machines are eventually set out to pasture and no longer receive the same updates to the driver as the more current machines. I have no objection to this, just pointing out an inaccuracy.

It eventually evolved much closer to my description.
Old series (M V X) had the old generation architecture (CPU, memory etc). Now all new series (since 2007) are using the external resource architecture. VLS series (VLS despktop and platforms) even have a common CPU board. PLS has a different CPU with separate drivers for X, Y and Z, but again they do not cost much as well. The bottom line is that all ULS machines use the same architecture and same control software which is also the printer driver. You should know how much progress has been made over the years. Your own PLS machine actually received a dozen or two of significant enhancements without moving a screwdriver or paying for new hardware.

I may upgrade the PC, but I am in no way upgrading the laser. I know this is the line ULS likes to use when comparing themselves against competitors that use on-board memory (e.g., Epilog), but it's simply marketing and no more. My laser does not any work faster with a top of the line computer than if I was using Windows 2000 on an old Pentium. It either works or it doesn't. Sure, an on-board memory chip might fail on a laser's motherboard, but so could any of the other chips... it's simply a repair that needs to happen, and I can guarantee a SIMM will be cheaper to replace even 10-15 years from now than a failing motherboard in the laser. Again, I have no real complaint about on-board versus off-board memory, just pointing out a flaw in the argument.

It's not just marketing. You can now process much more complex graphics at faster pace, you can use it with new operating systems, and when you buy a new PC with 64 bits and faster memory, some of the tasks in preprocessing takes much les times.
Today I can already see that memory chips for old X series machine cost more.
Not to mention that old machine CPU boards could handle limited amount of memory and the user is limited in how much he can squeeze into his laser machine. Think of a very large area - greyscale bitmap. Whereas with ULS method, you are just indifferent. You keep storing as many jobs as you are willing to, you just click and it runs the machine (no delay while transferring job files to laser).
This is in my book enhancement of the laser system, without even opening a lid of the machine.

I am sure that owners of chinese machines are happy with their equipment as well. But when I see them using it, without debating on other aspects, I see that their workflow has too many steps and the overhead for creating and/or running a job (two different tasks, FWIW) is significantly higher.

Zvi

[Ross Moshinsky]

I don't pay much attention to Universal but that method sounds like a terrible idea. Streaming is one of the most inconsistent ways to deal with information. What if there is a signal interruption? What if the computer freezes? What if I start watching a HD movie and it starts eating computer resources?

The best thing about a laser with on-board memory is, once it's sent over, the computer could literally be disconnected and it wouldn't matter. Considering a laser is only doing one thing and my PC while idling is doing 60, I'd pick on board memory every time. I wouldn't touch a laser that streamed the information if you paid me. One crash while running an expensive award or something one of a kind and it could cost hundreds if not thousands of dollars.

[Zvi Grinberg]

We aren't talking about US labor. We are talking about Chinese labor. Not to mention, we are talking about companies that should have a couple software people on staff. Assuming they make about $15 an hour, I have a hard time believing it would take them more than 67 hours to complete. I bet Laserworks took 10x as much programming time and they give that away for free.

I'm afraid you are under estimating the task, and over estimating the potential quality of the outcome. These are ends that never meet.

Programmers's job is never done. You write 90% of the software in 10% of the time, and the rest 10% of the software takes forever.
If you hire 15$ hour programmers you will get 15$ value of software. If you allocate budget for 67 hours or 1000$to fulfill the job you would get exactly what you paid for.

No long term commitment, no vision, no dedication - is a waste of resources for such products.