Should we transition to a dedicated RIP (Onyx vs. Caldera)?

[Warmoth]

Looking for opinions on dedicated RIP software like Onyx and Caldera.

Our shop currently runs 5 Roland roll to roll printers, and 6 Graphtec plotters. We like Versaworks (free) and we like Cutting Master (free). We do a lot of print/cut jobs utilizing the Roland/Graphtec combo, and it's great.
I really don't WANT to change anything, but it seems that it will become necessary if we ever decide to get something other than Roland for printing. We're looking into flatbeds and it seems unavoidable.

So far I've only looked closer at Onyx and I honestly have a huge issue with their business model. Subscription, upgrade, upgrade, upgrade, upgrade. You have to pay extra for additional concurrent RIP processes - Seriously? I've reached out to their sales and was told it'd be "In the ballpark of $13k", most of which is a one-time pay, and "There would be an optional annual cost of roughly $1,800 if you would like support, version and driver updates, and a couple of additional subscription features for Thrive".

Coming from the Roland / Versaworks ecosystem (free), I just don't see how that much money makes sense at all. How does it translate? It seems as though everyone has accepted this huge cost - why?

Is Caldera any better? Does anyone feel passionately about recommending we switch to a dedicated RIP? What are we missing out on?!

Any thoughts would be appreciated. Thanks.

 

[BigNate]

... so are you an amateur or are you a professional? the freebies seem to be best for the amateurs... though occasionally you find a tinkerer who works professionally as a sign maker and wants to tinker with the freebies - Onyx is very good, (so is Flexi). Get a good subscription with enough rips (I have 4 and rarely wish there were more - for as many devices as you have, maybe 6?).

Then once you have Onyx, figure out how many more man-hours you have from the more efficient workflow - then calculate what you had been paying an operator to work without a rip -- I'll bet you gain more time than the $1800 subscription fee.

IF a RIP costs more, then pass the little bit on to the customer.... (if it costs less, just don't tell anyone and let the production worker crank out a few more paying jobs in the same time and take the extra profit to the bank!)

 

[Warmoth]

... so are you an amateur or are you a professional? the freebies seem to be best for the amateurs... though occasionally you find a tinkerer who works professionally as a sign maker and wants to tinker with the freebies - Onyx is very good, (so is Flexi). Get a good subscription with enough rips (I have 4 and rarely wish there were more - for as many devices as you have, maybe 6?).

Then once you have Onyx, figure out how many more man-hours you have from the more efficient workflow - then calculate what you had been paying an operator to work without a rip -- I'll bet you gain more time than the $1800 subscription fee.

IF a RIP costs more, then pass the little bit on to the customer.... (if it costs less, just don't tell anyone and let the production worker crank out a few more paying jobs in the same time and take the extra profit to the bank!)

I'd consider ourselves professionals, we've been at it for over 30 years.

I guess what I'm trying to figure out is how does Onyx affect efficiency? What exactly does Onyx offer? (their website is truly a convoluted vague mess).

It takes like 5 seconds to setup a job using Versaworks. I don't see how it could possibly be any easier or faster than that? I haven't spent time in other shops, so I am very open to enlightenment on this. If we're missing out, I love advice!

 

[victor bogdanov]

What exactly does Onyx offer?

Can run just about any cutter/printer combo.

VersaWorks is very good but can only run Roland equipment.

Get whatever flatbed printer you are looking at and use the recommended/bundled rip for the flatbed and then decide if you want to switch all the printers to that RIP. No need to make the switch ahead of time

 

[FrankW]

VersaWorks and Onyx can´t be compared. It is like comparing a typewriter with a computer.

 

[FrankW]

The same with Caldera

 

[White Haus]

Can run just about any cutter/printer combo.

VersaWorks is very good but can only run Roland equipment.

Get whatever flatbed printer you are looking at and use the recommended/bundled rip for the flatbed and then decide if you want to switch all the printers to that RIP. No need to make the switch ahead of time

Agreed, VW is actually amazing for a free, OEM RIP platform. Very easy to use and just works.

We transitioned from a 100% Roland/VW shop to running everything through Onyx a few years back. It is a big step and I understand the confusion/frustration, but once you get it you'll wonder why you didn't do it sooner.

Like Victor said, it allows you to print on ANY device and then cut on ANY other device. No other software, no complicated workflows etc.

My only warning/complaint is that I've never been able to actually run a cut only job out of Onyx CutServer. They have some sort of nonsense workflow with SVG files and it just never works. Why they can't just work around PDF files like everything else I'm not sure, but I digress. Not sure how much cut only jobs you're running on your Graphtecs but that might be something that you continue to do with Cutting Master.

Yes, it is expensive. But it is a whole other level of automation and tools that free software just doesn't offer. If you haven't yet, ask them to give you a live demo where you can ask all the questions.

I have no experience with Caldera although you don't see a lot of complaints around here about it. I'm assuming it is more expensive/better than Onyx based on the comments I've read but that's only an assumption. (Someone please correct me if I'm wrong regarding pricing vs. Onyx)

I could go on and on about all the features that Onyx has, but they're all listed on their website. Spend a couple of hours watching all their youtube videos where they walk through all the features.

Regarding the ongoing expense...yes, like everything else they will continue you squeeze you until the end of time, unfortunately. Personally I just upgrade every few years when old versions stop being supported or when new major features are released. It's a cost of doing business like everything else, and in the end should (theoretically) improved your shop's efficiency and throughput.

 

[Pauly]

If you think Onyx is expensive, don't look at caldera.
Caldera would cost double Onyx.

Caldera you need Mac OSX or Linux for it to run.

I have no exp with versaworks, but not do I want to. Personal, I'm surprised you run 5 rollands and haven't decided to replace them all with something better and less machines. More doesn't equal better.

The difference in short.
Your using cutting master to setup and create the cut file and versa to print.
With. Rip software, it's done all in 1.
While it's done all together, there's much more powerful tools, e.g in Caldera there's true shape nesting, ink savings, colour management. And so much more.

If your a proper print shop, investing into the right set of tools is key.
Your lack of want into pay is concerning.
Yes, paying for licences for multiple printers is normal. Each licence has a licence for another adobe rip engine, which the rip manufacturers have to also pay for.
Because of you deciding to use 5x rollands, you'll need to buy 5x printer licences for any rip software you use.

I'm not here to give business advice, but I'd consider moving away from 5 rollands and look at something like an Epson 80600 or 2.

 

[WildWestDesigns]

If you think Onyx is expensive, don't look at caldera.
Caldera would cost double Onyx.

I think too, would have to get the more expensive options to get the perpetual license.

Caldera you need Mac OSX or Linux for it to run.

Very easy and cheap to get a Linux box setup, only downside is that they only support very specific versions (although really, Linux(because it's just the kernel), is Linux, knowledge is the key, I've gotten programs packed in a deb file to work on Arch very, very easily (deb, tar, rpm files etc are just compressed archive packages like a zip file)), so there is that, but bare in mind, no official support if go that route. Maya is another big commercial program available on Linux and they tend to support the RHEL based distros (can get it to work on Fedora, but no official support). Just keep that in mind, but I'm started to digress too much there.

Your lack of want into pay is concerning.

As someone that despises subscription services, I can somewhat agree with this if it comes at the sacrifice of abilities. If one has to have certain abilities within a program and the lower tiers (whatever that may be) do not have it and still don't get the upper levels, that's a big no no. If the abilities are there, but just require a different workflow, that may be something else.

 

[BigNate]

I'd consider ourselves professionals, we've been at it for over 30 years.

I guess what I'm trying to figure out is how does Onyx affect efficiency? What exactly does Onyx offer? (their website is truly a convoluted vague mess).

It takes like 5 seconds to setup a job using Versaworks. I don't see how it could possibly be any easier or faster than that? I haven't spent time in other shops, so I am very open to enlightenment on this. If we're missing out, I love advice!

the workflow will be much smoother, the switching between layout, in-position, cutting, color matching functions.... just about anything from the same platform - no switching programs, no making sure you can move the needed file between programs... just take the artwork into Onyx, get it ready, then print/cut/whatever....

I find even on a simple poster order I save about a minute vs using our old system of fixing in Illi or PS then setting in InDesign and printing... print 60 posters in a day and get a free hour! (I am sure whatever widget you produce you will find similar boosts in efficiency when using Onyx.)

if it takes 2 seconds in Onyx to setup a job you will realize a 60% savings in time.... (and let's be honest, you have not setup a job in Versaworks in 5 seconds.... possibly imported with already programmed presets (took time to preset, right?))

 

[Warmoth]

Agreed, VW is actually amazing for a free, OEM RIP platform. Very easy to use and just works.

We transitioned from a 100% Roland/VW shop to running everything through Onyx a few years back. It is a big step and I understand the confusion/frustration, but once you get it you'll wonder why you didn't do it sooner.

Like Victor said, it allows you to print on ANY device and then cut on ANY other device. No other software, no complicated workflows etc.

My only warning/complaint is that I've never been able to actually run a cut only job out of Onyx CutServer. They have some sort of nonsense workflow with SVG files and it just never works. Why they can't just work around PDF files like everything else I'm not sure, but I digress. Not sure how much cut only jobs you're running on your Graphtecs but that might be something that you continue to do with Cutting Master.

Yes, it is expensive. But it is a whole other level of automation and tools that free software just doesn't offer. If you haven't yet, ask them to give you a live demo where you can ask all the questions.

I have no experience with Caldera although you don't see a lot of complaints around here about it. I'm assuming it is more expensive/better than Onyx based on the comments I've read but that's only an assumption. (Someone please correct me if I'm wrong regarding pricing vs. Onyx)

I could go on and on about all the features that Onyx has, but they're all listed on their website. Spend a couple of hours watching all their youtube videos where they walk through all the features.

Regarding the ongoing expense...yes, like everything else they will continue you squeeze you until the end of time, unfortunately. Personally I just upgrade every few years when old versions stop being supported or when new major features are released. It's a cost of doing business like everything else, and in the end should (theoretically) improved your shop's efficiency and throughput.

Good to hear from someone who's made that same switch. I've watched some demos to try and get a better understanding. I see a few things so far that look kind of neat, but nothing completely game changing (for our particular output). I am still looking though, I am very open to change. Oh and we do quite a lot of cut only jobs.

Regarding the cost, there's nothing wrong with spending money to make money. I joked in another thread that we're cheap, but I completely agree with you - do it when it's necessary, or beneficial for the business.

 

[BigNate]

Good to hear from someone who's made that same switch. I've watched some demos to try and get a better understanding. I see a few things so far that look kind of neat, but nothing completely game changing (for our particular output). I am still looking though, I am very open to change. Oh and we do quite a lot of cut only jobs.

Regarding the cost, there's nothing wrong with spending money to make money. I joked in another thread that we're cheap, but I completely agree with you - do it when it's necessary, or beneficial for the business.

we setup cut-only jobs through Onyx all the time - just treat it like a print job and the cut file is created the same. Now need to actually print. pm if you need more details.

 

[Pauly]

I think too, would have to get the more expensive options to get the perpetual license.

Very easy and cheap to get a Linux box setup, only downside is that they only support very specific versions (although really, Linux(because it's just the kernel), is Linux, knowledge is the key, I've gotten programs packed in a deb file to work on Arch very, very easily (deb, tar, rpm files etc are just compressed archive packages like a zip file)), so there is that, but bare in mind, no official support if go that route. Maya is another big commercial program available on Linux and they tend to support the RHEL based distros (can get it to work on Fedora, but no official support). Just keep that in mind, but I'm started to digress too much there.


As someone that despises subscription services, I can somewhat agree with this if it comes at the sacrifice of abilities. If one has to have certain abilities within a program and the lower tiers (whatever that may be) do not have it and still don't get the upper levels, that's a big no no. If the abilities are there, but just require a different workflow, that may be something else.

If you're going to run Caldera, It's recommended to use Ubuntu.
But yes, you're correct, you could get it running on other distorts if you tried to. but it wont be supported.

The RIP subscriptions is for 2 things
1, Support.
2, Updates.
3* Features they want to put behind a pay wall.

Unless you mean the actual subscription service like Onyx Go. Which im not a fan of. only good if you cannot afford the costs of a perpetual licence.

 

[WildWestDesigns]

If you're going to run Caldera, It's recommended to use Ubuntu.
But yes, you're correct, you could get it running on other distorts if you tried to. but it wont be supported.

Yes. I only mention that because people seem to think that if a program packed as a .deb file (or .rpm etc) that it's locked in to a specific distro and it really isn't. Now, they make it that convenient for those that run that particular distro as that narrows down support issues, so there is a reason for them to steer people to those specific distros. It used to be Debian and/or Ubuntu (which Ubuntu is a derivative of Debian, just Debian isn't as "modern" in it's libs), but that could have changed.

The RIP subscriptions is for 2 things
1, Support.
2, Updates.
3* Features they want to put behind a pay wall.

No, I was gearing more to features.

Updates is a double edge sword. Not only in the stability part (oh the irony) as any change to a system is going to bring about a very good chance of instability, but also features can be deprecated/removed with the next version. Not so much of an issue if one doesn't use said features, but never know when one will get hit with that issue of always being on the latest and "greatest". Support in theory could be good. one of those things that is "better to have it and not need it compared to needing it and not having it".

I am also one that doesn't not like production computers connected to the WAN. LAN...sure, but not the WAN. I'm not liking that trend. We have just come full circle and back to no longer having PCs (I'm old school, "PC" is "personal computer", not really Windows specific which most think that it is now), but we are going back to the ole mainframe/terminal paradigm (for those that are old enough to remember that), just with a twist.

 

[Warmoth]

I have no exp with versaworks, but not do I want to. Personal, I'm surprised you run 5 rollands and haven't decided to replace them all with something better and less machines. More doesn't equal better.

The difference in short.
Your using cutting master to setup and create the cut file and versa to print.
With. Rip software, it's done all in 1.
While it's done all together, there's much more powerful tools, e.g in Caldera there's true shape nesting, ink savings, colour management. And so much more.

If your a proper print shop, investing into the right set of tools is key.
Your lack of want into pay is concerning.

Care to elaborate on your dislike for Roland? They've been good to us for many years. One of them is at least 12 years old, and it's been printing it's butt off nonstop. They're all extremely reliable, capable of great color, and pretty quick. If you say you have no experience with Versaworks, I imagine you've never used a Roland personally?

On some occasions we print and cut on one machine (with or without laminating). In these cases Versaworks handles the full setup. I have to say it quite literally takes about 5-10 seconds after importing the file to start printing. Perhaps some users aren't aware of the multiple Queues available, much like Onyx Quicksets (I think they're called), they have independent configurable defaults. They're even linked to network sharable hot folders. All I have to do is change the quantity for the job in most cases.

In cases where a Graphtec is utilized for cutting, we setup the array of graphics for the proper count and spacing (handled with macros or a very handy plugin called "eCut") directly in CorelDraw. A process that I'd say takes between 20-30 seconds? We simply export the sheet to be printed. As mentioned before, it takes mere seconds to start the print. When it's time to cut, select the cut layer in CorelDraw, send to Plotter. It's really very quick. If 20-30 sounds like a lot, the bottleneck in this production is truly not having enough printers. The files pile up waiting to be printed.

If semantics affects your opinion, we aren't a print shop, we are a sign shop. We do almost no paper printing at all. And I for one love investments. Really smart ones; like a Roland that's been paid off for over a decade that's still printing money everyday.
And since when did being financially diligent become a bad thing? If anything, your want to spend indiscriminately should be concerning.

 

[Pauly]

Care to elaborate on your dislike for Roland? They've been good to us for many years. One of them is at least 12 years old, and it's been printing it's butt off nonstop. They're all extremely reliable, capable of great color, and pretty quick. If you say you have no experience with Versaworks, I imagine you've never used a Roland personally?

On some occasions we print and cut on one machine (with or without laminating). In these cases Versaworks handles the full setup. I have to say it quite literally takes about 5-10 seconds after importing the file to start printing. Perhaps some users aren't aware of the multiple Queues available, much like Onyx Quicksets (I think they're called), they have independent configurable defaults. They're even linked to network sharable hot folders. All I have to do is change the quantity for the job in most cases.

In cases where a Graphtec is utilized for cutting, we setup the array of graphics for the proper count and spacing (handled with macros or a very handy plugin called "eCut") directly in CorelDraw. A process that I'd say takes between 20-30 seconds? We simply export the sheet to be printed. As mentioned before, it takes mere seconds to start the print. When it's time to cut, select the cut layer in CorelDraw, send to Plotter. It's really very quick. If 20-30 sounds like a lot, the bottleneck in this production is truly not having enough printers. The files pile up waiting to be printed.

If semantics affects your opinion, we aren't a print shop, we are a sign shop. We do almost no paper printing at all. And I for one love investments. Really smart ones; like a Roland that's been paid off for over a decade that's still printing money everyday.
And since when did being financially diligent become a bad thing? If anything, your want to spend indiscriminately should be concerning.

No dislike for Rolland, they're just an entry level machine. Same with their software. It's designed to do minimal tasks.

Your at 5 machines.. wouldn't you rather have 2 machines with the same capacity of the 5 you already have?
You factor in things like space, operational costs, electricity usage, keeping consumable for 5 machines vs 2 machines.

I had a choice to buy 4x fc9000s or 2x summa S3 160tcs.
I got the 2 summas. They are faster then the 4x fc9000s. They take up less space. It costs less to run 2 units vs 4. And in don't need to pay for someone to manage 4 of them.
Also because their faster, it takes less time for the job to pass though the machine to the next process.

Your process may not feel long, but once you've done all your tasks with 1 click rather than 5 clicks. It's quicker.
E.g. die cut stickers In caldera for example (as that's our current RIP)
I can open a file (print ready)
Tell it I need 100 copies. Add more bleed because I want 5mm bleed and the current 2mm is not enough.
Well have caldera set to nest contours to minimise space and maximise media usage.
We'll also have it set to use 500mm lengths at a time so handling the die cut stickers in a sheet easily.
Happy with the settings? Save it as a quick set.
Next time you do the same type of job, you just select the quickest and watch it get pushed to a printer.
(This is just basic stuff, you can do a lot more )
You can sent 20 jobs with the same requirements at once. 1 click.

Being financial diligent is a good thing, but understand what to invest in is something else. We spend a lot on software per year. I look at it this way. If a piece of software can save me time and money, we'll use it.
An example is our nesting software costs something like 4k per year. I've only had it for less than 6 months, and it's saved us over 10k of stock already.
Our automation, our web2print, etc all ads up. It's not cheap and not for the faint of heart, but without it, we would not be able to handle the 200-300 jobs we do per day.
It would cost me more in staff to do it manually. So yes, i won't hesitate to spend, if it saves time.
Even when where doing 10 jobs per day. Anything that saved time, I did.

Basically, I'm not saying to go spend 15k on caldera or remove all your rollands and replace them with something else.
All I'm saying is think about the workflow and think about what will save you time and money.
You mentioned if the bottleneck is not having enough printers, you should think about.. do I really need a 6th? Do I have space for 6? If so, should I get something faster and cheaper to run so a 7th is further away? 12 years is getting old, might need to retire it, so then I'll need to replace that with another or will the faster one replace that at the same time?

Same goes with your flatbed thoughts.
Maybe stretch your budget for a HP R1000. (Or a different brand)
It's a hybrid so you can move the table/rollers and use it as a roll printer.
You've just gained capacity as it's a R2R and extra capabilities with flatbed work.

 

[victor bogdanov]

Care to elaborate on your dislike for Roland? They've been good to us for many years. One of them is at least 12 years old, and it's been printing it's butt off nonstop. They're all extremely reliable, capable of great color, and pretty quick. If you say you have no experience with Versaworks, I imagine you've never used a Roland personally?

I had 5 Rolands at one point and changed them out for a Colorado and Epson Resin. The Colorado alone is about is faster than the 5 Rolands were and it is much easier to run 1 printer vs 5. (Rolands could do a 60" x 150ft roll in 6+hrs each), on the Colorado I can do a roll in a little over an hour.

The rolands are great printers for starting out/low volume but when you need 3+ to keep up it is a good idea to level up to the next tier of printer. Epson S60600 would be a good upgrade

 

[Warmoth]

No dislike for Rolland, they're just an entry level machine. Same with their software. It's designed to do minimal tasks.

Your at 5 machines.. wouldn't you rather have 2 machines with the same capacity of the 5 you already have?
You factor in things like space, operational costs, electricity usage, keeping consumable for 5 machines vs 2 machines.

I had a choice to buy 4x fc9000s or 2x summa S3 160tcs.
I got the 2 summas. They are faster then the 4x fc9000s. They take up less space. It costs less to run 2 units vs 4. And in don't need to pay for someone to manage 4 of them.
Also because their faster, it takes less time for the job to pass though the machine to the next process.

Your process may not feel long, but once you've done all your tasks with 1 click rather than 5 clicks. It's quicker.
E.g. die cut stickers In caldera for example (as that's our current RIP)
I can open a file (print ready)
Tell it I need 100 copies. Add more bleed because I want 5mm bleed and the current 2mm is not enough.
Well have caldera set to nest contours to minimise space and maximise media usage.
We'll also have it set to use 500mm lengths at a time so handling the die cut stickers in a sheet easily.
Happy with the settings? Save it as a quick set.
Next time you do the same type of job, you just select the quickest and watch it get pushed to a printer.
(This is just basic stuff, you can do a lot more )
You can sent 20 jobs with the same requirements at once. 1 click.

Being financial diligent is a good thing, but understand what to invest in is something else. We spend a lot on software per year. I look at it this way. If a piece of software can save me time and money, we'll use it.
An example is our nesting software costs something like 4k per year. I've only had it for less than 6 months, and it's saved us over 10k of stock already.
Our automation, our web2print, etc all ads up. It's not cheap and not for the faint of heart, but without it, we would not be able to handle the 200-300 jobs we do per day.
It would cost me more in staff to do it manually. So yes, i won't hesitate to spend, if it saves time.
Even when where doing 10 jobs per day. Anything that saved time, I did.

Basically, I'm not saying to go spend 15k on caldera or remove all your rollands and replace them with something else.
All I'm saying is think about the workflow and think about what will save you time and money.
You mentioned if the bottleneck is not having enough printers, you should think about.. do I really need a 6th? Do I have space for 6? If so, should I get something faster and cheaper to run so a 7th is further away? 12 years is getting old, might need to retire it, so then I'll need to replace that with another or will the faster one replace that at the same time?

Same goes with your flatbed thoughts.
Maybe stretch your budget for a HP R1000. (Or a different brand)
It's a hybrid so you can move the table/rollers and use it as a roll printer.
You've just gained capacity as it's a R2R and extra capabilities with flatbed work.

I see where you're coming from in some regards, but I think you may have an outdated bias against Roland. I certainly don't think they are the 'De Facto' printer, but I checked the specs on the Epson you suggested and it's production speed shows 310 sq ft / hour, compared to the Roland's spec sheet of 327 sq ft / hour (I'm sure real experiences may vary). The Epson had a slight edge on DPI.

I do find it convenient having several printers though, it reduces media juggling. Each printer can print anything when needed, but we have designated default media types for each of them. So I'm not really sure if I'd prefer having just 2 printers - they'd have to be significantly faster, I suppose. (The Colorado looks fast).

Also I'm curious now, you had previously cited nesting capabilities as a selling point for Caldera, are you now saying that you have additional software for nesting as well?

For the flatbed, we're now looking very seriously at a Vanguard VR5D-E (comes with Onyx) We like the ability to add heads. The price is nice, and local techs on call. The plan is to pull the trigger on Monday, unless we can talk ourselves out of it before then.

 

[victor bogdanov]

I do find it convenient having several printers though, it reduces media juggling. Each printer can print anything when needed, but we have designated default media types for each of them. So I'm not really sure if I'd prefer having just 2 printers - they'd have to be significantly faster, I suppose. (The Colorado looks fast).

Another advantage of moving up to the next level of printers is ink cost. Epson with bulk in and Colorado M5 will be significantly cheaper ink per sq ft than roland. My ink cost just about got cut in half when I switched from Roland to Colorado and the savings on ink pay for the Colorado. Bulk ink Epsons also offer good savings

 

[Pauly]

I see where you're coming from in some regards, but I think you may have an outdated bias against Roland. I certainly don't think they are the 'De Facto' printer, but I checked the specs on the Epson you suggested and it's production speed shows 310 sq ft / hour, compared to the Roland's spec sheet of 327 sq ft / hour (I'm sure real experiences may vary). The Epson had a slight edge on DPI.

I do find it convenient having several printers though, it reduces media juggling. Each printer can print anything when needed, but we have designated default media types for each of them. So I'm not really sure if I'd prefer having just 2 printers - they'd have to be significantly faster, I suppose. (The Colorado looks fast).

Also I'm curious now, you had previously cited nesting capabilities as a selling point for Caldera, are you now saying that you have additional software for nesting as well?

For the flatbed, we're now looking very seriously at a Vanguard VR5D-E (comes with Onyx) We like the ability to add heads. The price is nice, and local techs on call. The plan is to pull the trigger on Monday, unless we can talk ourselves out of it before then.

I only used Epson as an example as most will move from Roland to Epson. Im not sure which Rolands you're using.
The Colorado M3 and M5 are on another level. and yes they would replace all Rolands with 1 unit.
And as Victor mentioned, Ink cost is less.
Roland is an entry level brand. At the price point they sit at, it's classified as entry level. It's not a bad thing. Same with Mutoh, Mimaki and HP, although HP has both entry and production level stuff.

Yes, Caldera has great nesting capabilities. But yes, we use additional software to nest.
Reason being is we use Enfocus Switch for our automation combined with other software. All our jobs will be prepressed and nested then sent to caldera. If im going to submit something to print manually, then we'll use calderas nesting features.

This is some advice for a flatbed printer.
Find the ones you're interested in.
Go get demos of them all
Get samples, get your own files printed. Compare them with each other
Then make your decision..

Dont buy one with out getting a demo.

Also the ability to add heads is not a selling point.
Most techs will tell you its not a good idea due to having a new row next to an old row. You will get density banding and other weird stuff.