New Phrozen Printer "Sonic Mighty Revo"

[Peachymike]

THAT looks like we're already there.  I don't see any layer lines...and neither does the close-up shot.

EXCELLENT!

Cheerio!
Bob Gilmore

Thanks Bob!!

[glakedylan]

Plus, the rivet detail is amazing!

[Lemosteam]

Man you guys are a tough audience! OK I thought to compare apples to apples I put a light coat of Tamiya primer on it. Took the picture with my wifes phone which has a much better camera, at a similar angle. Lighting is a little dark on the smoke box front. The boiler is almost exactly 0.5 inches across. The whole boiler looks like this, nary a line to be seen. Proof is in the pudding so to speak... (Attachment Link)
Mike

Everyone always takes my commentary as personal criticism of their work. It's not, and yes I can still make out layer lines in all of your images so far.  It is merely stating facts regarding the fallacy of replacing a highly polished mold, with a process tech riddled with potential error states that can occur, no matter what the user attempts, particularly where repeatability is concerned. I also realize most folks are not using this as a manufacturing process, save for nebulous items like detail parts for scenery (in the MRR world anyway).

I have always agreed that there is a point of diminishing returns for this tech.  The smaller the voxel, the more time is involved.  The more time you have to print, the higher the possibility of other error states to occur. 

Secondly Robert, all of the various natural things you mention are inherently, and naturally round or round-ish, even a fiber is typically a round cross section that tapers at each end of its length, lending to the natural curvature and smoothness of the things we see and touch every day (ignoring the microscopic).

Voxels, even rounded on the hard corners via anti-aliasing, etc., are unnatural in that sense, hence the reason we can see and feel the edges, under magnification or not. I also think that viscosity of the resin comes into play, filling valleys that cannot be completely cleaned and remaining after the curing process (a good thing to some extent).

Layered SLA depends on three things.  The accuracy of the light "edges" and why I think DLP helps, the layer height accuracy of the machine (and to Robert's point even the 18 micron nominal will have a tolerance range on it of +/- ? microns), and the STABILTY of the object being printed, the most commonly ignored and difficult to control error state.

The taller the print, the more stability is a factor in layer lines and I think this is why many folks try to print flat or with curved shapes parallel to X-Y plane.

Evidence of this is present, obvious to me, in the last image.  Again, YES I think this looks awesome compared to other attempts I have seen and as Robert points out much closer to the human eye perspective, so don't get offended'.

These are clearly lines representing the angle at which the boiler was printed, and guessing that the smokebox was the last portion to print.

Would also love to see your model where the supports were removed and how clean that area is (pink elephant in the room, also not a criticism, which I know is COMPLETELY out of anyone's control).

[ Guests cannot view attachments ]

[robert3985]

I think the point here is that there will always be layer lines in resin 3D printing, and you'll be able to perceive them when you enlarge a sharp photo of your model by 500 to 1750%...or beyond.

But, from a practical standpoint...and from a standpoint of "normal" human eye visual acuity, layer lines like on Mike's print are practically invisible, especially after painting & weathering.

Personally, I'd be plenty happy with these results on any model I printed for my N-scale world.

As for supports...gotta put 'em where they are least visible and easy to clean up which MSLA 3D printing allows as opposed to removal from sprues, flash, sink holes and mold relief on injection molded models.

Cheerio!
Bob Gilmore

[peteski]

I think the point here is that there will always be layer lines in resin 3D printing, and you'll be able to perceive them when you enlarge a sharp photo of your model by 500 to 1750%...or beyond.

But, from a practical standpoint...and from a standpoint of "normal" human eye visual acuity, layer lines like on Mike's print are practically invisible, especially after painting & weathering.

It is not just the human eye we have to consider.  Many model railroaders taking close-up photos of their models (as you in many thread see on this forum). Those photos will expose layering. But I also agree that at this point in time many SLA hobby-grade printers have resolution high enough that that with proper painting techniques, models look quite good.

This entire discussion started from Mike's comment that the smaller size voxels do not make an appreciable difference.  John then introduced some other factors that affect the print quality of DLP printers.  I was just discussing the visual surface finish of complex curved surfaces as related to voxel size. I suppose we all have different standards.

[Lemosteam]

I didn't argue that point, Robert.  In fact I said I agree.

Mold processes can be managed and easily hidden, not so much for supports as they must touch nearly every unsupported feature hanging out there coming off the FEP, not to mention the resin coagulation that inevitably occurs at the tip of each support. 

Supports affect at least one viewing angle, guaranteed.

All I am trying to say is that resin printing is just another tool in the toolbox, not a replacement for injection molding. Do they get close enough, sure.

[Peachymike]

I think everyone is missing the whole point, the original question was about “upgrading” an already excellent machine to one that provided a dubious increase resolution, at a cost of $999. My point was that with good technique and a little experimentation excellent results are possible, mostly negating the advantage of the slight increase in resolution. This will never replace injection moulding even if it does equal or surpass it. But for $300 and a $25 litre of resin I can make some pretty kick butt models that aren’t available any other way. Remember too the machine is only part of the equation, the current resins are at the limit of what they can reproduce, imho. Files, supporting, settings , etc, all play a part. Bty the supported section of the boiler is the bottom edge, you can see a few of the attachments as this is a raw print, other than the primer which was to highlight and not hide.
I’m pretty happy and impressed with what I can do with my current equipment. When it wears out or there is a really significant change I will undoubtedly purchase that next generation of machine. I like to share what I have learned and hope that others would do the same, or build on it, rather than a p-ssing match of what’s wrong with it. As Peteski noted it all depends on your standards. This is more than good enough for me, and pretty much everyone I know so I am good with it. Hope your endeavours will yield the better results you require. 
Mike

[Chris333]

I have plenty photos that you can barely see any lines:


Just posted the last pic to say if you look close enough (closer than eyes?) there are lines.

Certain parts may achieve injection plastic levels. But until they fix the FEP tension on different surfaces we got a long way to go. I think it could be done with software that can see a thick area of a model right next to a thin area. Split those layers into 2 exposures that mask out the parts needed so that in whole the exposure it even.

[Lemosteam]

I have plenty photos that you can barely see any lines:


Just posted the last pic to say if you look close enough (closer than eyes?) there are lines.

Certain parts may achieve injection plastic levels. But until they fix the FEP tension on different surfaces we got a long way to go. I think it could be done with software that can see a thick area of a model right next to a thin area. Split those layers into 2 exposures that mask out the parts needed so that in whole the exposure it even.

Sorry Chris, side of cylinder, from the centerline down.   

[Chris333]

Sorry Chris, side of cylinder, from the centerline down.   

You shut your mouth when you're talking to me.

[wazzou]

I think it’s completely disingenuous to make comparisons between 3D printing and injection molding.
Let’s just start with cost and then the intent behind using the two mediums.

[Peachymike]

Everyone always takes my commentary as personal criticism of their work. It's not, and yes I can still make out layer lines in all of your images so far.  It is merely stating facts regarding the fallacy of replacing a highly polished mold, with a process tech riddled with potential error states that can occur, no matter what the user attempts, particularly where repeatability is concerned. I also realize most folks are not using this as a manufacturing process, save for nebulous items like detail parts for scenery (in the MRR world anyway).

I have always agreed that there is a point of diminishing returns for this tech.  The smaller the voxel, the more time is involved.  The more time you have to print, the higher the possibility of other error states to occur. 

Secondly Robert, all of the various natural things you mention are inherently, and naturally round or round-ish, even a fiber is typically a round cross section that tapers at each end of its length, lending to the natural curvature and smoothness of the things we see and touch every day (ignoring the microscopic).

Voxels, even rounded on the hard corners via anti-aliasing, etc., are unnatural in that sense, hence the reason we can see and feel the edges, under magnification or not. I also think that viscosity of the resin comes into play, filling valleys that cannot be completely cleaned and remaining after the curing process (a good thing to some extent).

Layered SLA depends on three things.  The accuracy of the light "edges" and why I think DLP helps, the layer height accuracy of the machine (and to Robert's point even the 18 micron nominal will have a tolerance range on it of +/- ? microns), and the STABILTY of the object being printed, the most commonly ignored and difficult to control error state.

The taller the print, the more stability is a factor in layer lines and I think this is why many folks try to print flat or with curved shapes parallel to X-Y plane.

Evidence of this is present, obvious to me, in the last image.  Again, YES I think this looks awesome compared to other attempts I have seen and as Robert points out much closer to the human eye perspective, so don't get offended'.

These are clearly lines representing the angle at which the boiler was printed, and guessing that the smokebox was the last portion to print.

Would also love to see your model where the supports were removed and how clean that area is (pink elephant in the room, also not a criticism, which I know is COMPLETELY out of anyone's control).

(Attachment Link)

I wasn't going to add anymore to this thread as it is sort of a standoff of opinions at this point, but I thought that for those that are novices it might be of interest to explain the nature of the these lines and methods to lessen them. These are not layer lines in the sense of the stepping cause by the ratio of the layer height to the pixel size, rather these in particular are a layer shift caused by increased suction on the FEP (which Chris333 so correctly identified as another issue). If they were layer lines  from the former then they would be consistent all the way down the side of boiler on that part of the curve. You can see this on Chris's excellent close up that shows these line evenly spaced in relation to one another.  IF you follow the lines up and around the boiler (not seen in the photo) on the above photo from the highlighted lines they align with the start of the sand dome on the angle where it started to print. The guy in the mirror screwed up, I made at least three mistakes. One was I didn't hollow out the inside of the dome top decrease cross sectional area on the FEP when the dome started to print. This  caused it to pull the print slightly out of alignment as it releases. The second was I should have put more support in that area to lessen the effect. The third was as I said in an earlier post, was a sloppy FEP. The FEP is probably the most flexible part of the system and the more area that is exposed at one time the greater the suction forces will be. This also has implications as to how full the plate is, I can do 20 of these at a time if I fill the build plate. This increases the suction across the FEP and since it is flexible it tends to be worse in the center than the edges. And a fourth was I could have added a "pause" in the time between layers which would have allowed the print a little time to rebound to its original position. I created the situation that lead to these lines by less than perfect technique. Analyzing why they happen can lead us to solutions that can lessen them even if we can't totally eliminate them. Even the highest resolution machines will not eliminate this, since it is not a function of the resolution, but rather related to the FEP
I hope that some will find it useful in their endeavours in 3d printing and provide some considerations for what they require to produce results that are acceptable to them.
Mike

[Lemosteam]

All of these points are valid, as are mine. There is no difference of opinion here. Everyone here knows I support the value of any type of 3D printing.

These novices who might read these posts must also realize that this is NOT a rigid science or a simple math equation. It is more of a witchcraft, and some users are better at honing the process than others.

There are so many variables that can negatively affect a print, that many just give up. Facebook groups are teeming with folks that are so frustrated trying to perfect a process that has moving targets. Room temp, resin temp, resin viscosity, Fep material, leveling, leakage, cleaning and curing methods, and I won’t even manti9n the miryiad of slicer settings, etc., etc. I’ve seen posts of people running print farms like “I’ve been printing this same file on three machines for five months straight.” , all things the same, no matenance changes or issues, and one of the three machines gets a delamination issue, or total goop fail, or won’t stick to the plate, etc.

3D printing has got to have the worst repeatability numbers ever, just due to the number of potential failure modes. I wonder what the failure probability would be, @GaryHinshaw after taking ALL of the variables into account.

It’s much like a perfect golf swing, you have it or you don’t. @Peachymike , I would suggest that you have it. I don’t. I go to bed with a print running and I pray all night that I will wake up to a decent, not even perfect, print, and not something that I have to peel off the FEP.

This is NOT an easy or trouble free process, I.e. “just do this”, or “just do that” is almost never the answer. Machine to machine, resin to resin, environment to environment, person to person, the probability of failure is very high.

I know my posts about resin printing sound bleak, but I feel that folks should also know all of the benefits and warts together.

What I find really amusing is that that other printing house gets blasted for their product output and surface quality all the time, no no one likes to show or admit to resin printing flaws that are in many cases worse, especially when one takes into account support removal and the resulting surface that it leaves.

@Peachymike , may I ask you how many tries it too to get that model to print that well. I am not being sarcastic here, but genuinely curious. I have one model that it took four vat-fulls of resin to get a reasonably satisfactory result. It

To each person, every tool has its purpose, and each persons purpose for using that tool may be different, and that is why, to me, every tool can benefit each modeler differently. After all, most of us are in fact modelers first. So let’s just sand off that Shapeways FUD, and smooth those pock marked support areas, and continue modeling, both electronically, and by scratch.

I just get weird when resin printing is made to sound rosy. It’s not, and we can all admit our tribulations. I even regret not waiting for a large volume DLP now.

[Peachymike]

All of these points are valid, as are mine. There is no difference of opinion here. Everyone here knows I support the value of any type of 3D printing.

These novices who might read these posts must also realize that this is NOT a rigid science or a simple math equation. It is more of a witchcraft, and some users are better at honing the process than others.

There are so many variables that can negatively affect a print, that many just give up. Facebook groups are teeming with folks that are so frustrated trying to perfect a process that has moving targets. Room temp, resin temp, resin viscosity, Fep material, leveling, leakage, cleaning and curing methods, and I won’t even manti9n the miryiad of slicer settings, etc., etc. I’ve seen posts of people running print farms like “I’ve been printing this same file on three machines for five months straight.” , all things the same, no matenance changes or issues, and one of the three machines gets a delamination issue, or total goop fail, or won’t stick to the plate, etc.

3D printing has got to have the worst repeatability numbers ever, just due to the number of potential failure modes. I wonder what the failure probability would be, @GaryHinshaw after taking ALL of the variables into account.

It’s much like a perfect golf swing, you have it or you don’t. @Peachymike , I would suggest that you have it. I don’t. I go to bed with a print running and I pray all night that I will wake up to a decent, not even perfect, print, and not something that I have to peel off the FEP.

This is NOT an easy or trouble free process, I.e. “just do this”, or “just do that” is almost never the answer. Machine to machine, resin to resin, environment to environment, person to person, the probability of failure is very high.

I know my posts about resin printing sound bleak, but I feel that folks should also know all of the benefits and warts together.

What I find really amusing is that that other printing house gets blasted for their product output and surface quality all the time, no no one likes to show or admit to resin printing flaws that are in many cases worse, especially when one takes into account support removal and the resulting surface that it leaves.

@Peachymike , may I ask you how many tries it too to get that model to print that well. I am not being sarcastic here, but genuinely curious. I have one model that it took four vat-fulls of resin to get a reasonably satisfactory result. It

To each person, every tool has its purpose, and each persons purpose for using that tool may be different, and that is why, to me, every tool can benefit each modeler differently. After all, most of us are in fact modelers first. So let’s just sand off that Shapeways FUD, and smooth those pock marked support areas, and continue modeling, both electronically, and by scratch.

I just get weird when resin printing is made to sound rosy. It’s not, and we can all admit our tribulations. I even regret not waiting for a large volume DLP now.

I more than agree with you, there are a LOT of pitfalls. It takes time to learn and its not just one discipline. As I said before, drawings, slicing, setup, support, resin choice, conditions that you print under, and on and on. I have had my share of failures, but I try to learn from them. And share those results in a positive light. Success breeds success. Start out simple with small projects and build on that. As you noted there are a ton of people that have problems, but there is a cause and effect and analyzing those failures helps me to get better. I have learned far more from failing and trying again than being lucky and getting a good print. I'm an old truck mechanic, if I can learn it, so can others. I love to tinker, and am somewhat tenacious so those things help me.
I do test prints all the time, one of the reasons I have several printers. All my printers are ELEGOO, they have been good to me as has the company. I test on the small printer, to get it to where I want it and then move to the bigger printer. I use mostly Siraya resins as I have them dialed into my printers fairly well. I keep the environment as close to the same as I can, try to print at the same ambient temps. The devil is in the details... 
To answer your question, when I started I did a lot of prints to get things right, because I had to figure it out. Now I would say 3 test prints average, and I usually have it to the point that it is acceptable. But that took a lot of time to learn, 5 years since I started now, and its been a journey. Anyone that thinks they can buy an STL, plunk it on the machine and voila, have a marvelous model is in for a big disappointment. You know that, and you know a lot about printing as is evidenced by wealth of information you have brought to this and other discussions. The fact that despite your problems you continue says a lot, you see the value. I love what I can do with this tech, and I only see it getting better, but no pain no gain. The learning curves are steep. And this is all about having fun, not driving yourself around the bend! That's why I try to share when I see an opportunity, If I can save someone a little aggravation, or inspire them to try again, who knows, they might be helping me out in a few years.
Because I try very hard to reduce my variables as much as I can, i get a fairly high rate of success NOW. But even with that a good 10% go straight into the trash. Then I break a few, or a dead pixel left a line on a surface, or any of the other myriad of problems that we can have rear there ugly heads. But I get 3/4 of what was on the plate as decent prints, and that is a wonderful thing. My glass is 3/4 full!
I agree with the DLP too, I think its the future, it will just take time to make it bigger and more affordable. And who knows what else is in the pipe. Things like the heated vats, roller lead screw and increased resolution are all going to help, making it easier for those that don't have the time or inclination to really dive in. No one wants to spend their hobby time on fruitless endeavors. And that is the crux of it, time.
I sure don't have all the answers but if we all work together and share, we all get better. That is my only intent. I sure learned a lot from forums like this and reading what other do, so maybe others can learn from what you and I and so many other have to share.
Mike

[Chris333]

When I bought my Sonic Mini 8Ks I was going to buy a DLP at the same time. Wanted one again, but got a Mars 4 ultra instead. Still want a DLP, but been paying off my house like crazy... Give me a few months 

Think I'm on my 10th print for a simple loco body right now. Everything I keep changing is to get rid of ripples from the FEP tension. A big problem was large body mounting lugs inside the body. I finally took them off and will print those separately to glue in place later. Hey it got rid of the ripples.