Drag and Drop Browser Seps

[AdvancedArtist]

I just want the art from the PS file if there is one and if not then oh damn I have to knock out the black background that will be more work than the sep.. LOL

[DDSol]

Thanks Dan.

I'm not saying these people don't know what they're talking about. Just that they don't know what they're talking about when it comes to my thing. I don't think either of those guys actually looked at it. I'm quite sure they put a lot of work into getting so experienced and that they know quite a bit of their stuff. Really, I'm sure. But they just can't say that what I'm making isn't possible. It's insane they started saying this right out of the gate, with not even a hint of actual investigation.

Anyway, this thing produces no seps yet, so I can't print tests using it, let alone demos.

[IntegrityShirts]

Now that we've established that you don't know what we're talking about and we don't know what you're talking about, maybe we can all take the time to explain what we're talking about without the abrasiveness?

Screen printers are sometimes very matter-of-fact without explanation.  There's a lot of "that won't work" that goes on here and it all has to be taken with a grain of salt. I see it every day here but I keep an open mind and don't necessarily believe what everyone types.  After all, this is the internet, and it stands to reason that most people on the internet are experts and don't type a lot of "in my opinion" and "I think" or "I believe".

I think the best disclaimer with ANY separations handed to a any Joe screen printer is: "your results may vary"

The hardest part of your job DDSol isn't the coding, it's accumulating all these opinions, then drawing conclusions defining concrete variables.

This thread reminds of....

http://embed.break.com/NDg4NzIx/ai/0/zi/0/ds/1/st/embed

[Sbrem]

Well, I've read through, and the info seems like a great idea, and though it apparently needs some thorough testing, I think we ought to give it some time to develop and see what happens. I like his enthusiasm, though not his exact delivery, but I think he truly is trying to make something pretty damn useful. A lot of know our way around, and are doubtful as to the claims, but let's keep this going, and when it's ready to give us files to print out and get onto screens, we'll have a great group of printers testing it. Remember to play nice...

Steve

[AdvancedArtist]

Now that we've established that you don't know what we're talking about and we don't know what you're talking about, maybe we can all take the time to explain what we're talking about without the abrasiveness?

Screen printers are sometimes very matter-of-fact without explanation.  There's a lot of "that won't work" that goes on here and it all has to be taken with a grain of salt. I see it every day here but I keep an open mind and don't necessarily believe what everyone types.  After all, this is the internet, and it stands to reason that most people on the internet are experts and don't type a lot of "in my opinion" and "I think" or "I believe".

I think the best disclaimer with ANY separations handed to a any Joe screen printer is: "your results may vary"

The hardest part of your job DDSol isn't the coding, it's accumulating all these opinions, then drawing conclusions defining concrete variables.

This thread reminds of....


http://embed.break.com/NDg4NzIx/ai/0/zi/0/ds/1/st/embed

ROFLOL I have talked with Han several times. He is a genius and very smart and knows what he is talking about on the science and math side of things.  And remember your just looking at a test a work in progress. But I am impressed and I can see the potential of something like this.. Innovation is coming Han will have some and so will others. But if we as a community can accept and work with that then things are going to look allot different in a short time.

[AdvancedArtist]

Again, it's not to put you or your product down. I have much respect for your products. It's a rare opportunity for all interested.

Dan I appreciate this post and I am all for it except one thing.. Steve would be so far out of his league he would be clueless and it would not be fair. Even this day new news relating to how we look at color has come forth...

http://www.huffingtonpost.com/2013/06/12/layer-human-eye-duas-layer-cornea_n_3427580.html?utm_hp_ref=mostpopular

Channels in PS are stone age right now and I can prove that mathematically and scientifically without a doubt. We never should have even been in those PS channels we were fools to even go there.

We have only scratched the surface of color in this Industry believe me. There are things I know I have been trying to figure out how to make simple and I will make it simple but that will take time as I have so many projects on my plate.

But Han has presented some new and very interesting things I did not try to hide that I exposed it I have no fear of reality, math or science.

[DDSol]

Hello. http://my.screenseps.com/seps.html I fixed it.

Because I did it wrong! But it's right, now.

Btw, I think that art should be in 6 colors, not 5. Well, the blue of the windows can otherwise not be made and it'll be more teal. Not ugly, just not the original .

And colors are not accurate. (Very accurate (Scary accurate)).

[AdvancedArtist]

While we are all debating here.. it would take me less than 5 minutes of video to forever change this ahhh the deep secrets and only 2 men on this planet of 6 billion have those secrets. Actually it has been here for a long time even before we had computers. It is primitive really and very basic just like HSB or HWB.

[DDSol]

I added a white base, and now it looks washed out. I don't like it. What's up? Use Less base? I have a weird formula for the white base, but it seems to work pretty well.

Let me know what you think. OH, btw, the white base has no "slot" yet and it's always on, unless you view the original. It's not a "sep" so to speak, so it demands special handling (still). I'll fix that later.

[GraphicDisorder]

The way Tom keeps "hyping" this reminds me of the talk of the invention that was going to change the world.  It was hyped for awhile by some really rich and powerful people before anyone even knew what it was.  This invention came out many many years ago and guess what.  It didn't change anything really.  Anyone remember what it was?
   
         
   
   
   
 
It was called a Segway. 

[aauusa]

As a small business shop with few employees so the few do the work of the many...dream of the day I can take an image and just drop it in a browser and it separates automatically for me.  and it does a 90% perfect job for me.   is it here now but i do see it on the horizon.  heck with the topic that is being talked about on this thread in the beginning would never been possible 5 years ago nor dreamt about.  so keep it up  I will also do some test with it this weekend I have some high end stuff I would like to see how it works.  I know it is a beta and no seps but should be fun to just see.

On a side note to AA.   I have  your simple seps 1.2 and the more I have used it it does give very good results.  It does have a great learning curve to it but the more I use it the easier it is so keep it up.  and the last job i used with it had a crappy jpg file in a customer supplied power point file.  was able to sep it very nicely.

[IntegrityShirts]

As a small business shop with few employees so the few do the work of the many...dream of the day I can take an image and just drop it in a browser and it separates automatically for me.  and it does a 90% perfect job for me.   is it here now but i do see it on the horizon.  heck with the topic that is being talked about on this thread in the beginning would never been possible 5 years ago nor dreamt about.  so keep it up  I will also do some test with it this weekend I have some high end stuff I would like to see how it works.  I know it is a beta and no seps but should be fun to just see.

On a side note to AA.   I have  your simple seps 1.2 and the more I have used it it does give very good results.  It does have a great learning curve to it but the more I use it the easier it is so keep it up.  and the last job i used with it had a crappy jpg file in a customer supplied power point file.  was able to sep it very nicely.

I think the 90% is a good point, at least for me.  I can use other sep programs and still spend the time to separate art and get to something that's 90% ok.  If I could just use the browser seps and they get me to the 90% without the other sep programs, it would save a ton of time.  For me, most of my customers aren't ordering quantities high enough to justify the cost of professional separations, but for now if it comes down to choosing, I'd still choose the human pro seps for piece of mind over the programmatic seps and here's why.

I 100% believe that it is possible for a computer to sep an image just as well if not better than a human given the same art to start with.  What hasn't really been touched on much here, is the color correction and bumps made to the art by the professional separator BEFORE seps are made.  Therein lies a good amount of the talent in my opinion.

I work with programmers in my day job and I make stylistic changes to interfaces and TINY tweaks to the appearance of some elements a lot.  The programmers all get sort of angry and ask, "Why did you change that, you said before you wanted X not Y?".  My answer is usually, "It LOOKS better now.", to which I get this glazed look of disbelief. They think that if I did it once a certain way before, then that's the only correct way of doing it in the future. To change it later on down the road means I don't know what I'm doing. A lot of programmers don't understand art, or the objective nature of design, to find both qualities in one person, programmer and artist, is truly a great thing.

[DDSol]

As a small business shop with few employees so the few do the work of the many...dream of the day I can take an image and just drop it in a browser and it separates automatically for me.  and it does a 90% perfect job for me.   is it here now but i do see it on the horizon.  heck with the topic that is being talked about on this thread in the beginning would never been possible 5 years ago nor dreamt about.  so keep it up  I will also do some test with it this weekend I have some high end stuff I would like to see how it works.  I know it is a beta and no seps but should be fun to just see.

On a side note to AA.   I have  your simple seps 1.2 and the more I have used it it does give very good results.  It does have a great learning curve to it but the more I use it the easier it is so keep it up.  and the last job i used with it had a crappy jpg file in a customer supplied power point file.  was able to sep it very nicely.

I think the 90% is a good point, at least for me.  I can use other sep programs and still spend the time to separate art and get to something that's 90% ok.  If I could just use the browser seps and they get me to the 90% without the other sep programs, it would save a ton of time.  For me, most of my customers aren't ordering quantities high enough to justify the cost of professional separations, but for now if it comes down to choosing, I'd still choose the human pro seps for piece of mind over the programmatic seps and here's why.

I 100% believe that it is possible for a computer to sep an image just as well if not better than a human given the same art to start with.  What hasn't really been touched on much here, is the color correction and bumps made to the art by the professional separator BEFORE seps are made.  Therein lies a good amount of the talent in my opinion.

I work with programmers in my day job and I make stylistic changes to interfaces and TINY tweaks to the appearance of some elements a lot.  The programmers all get sort of angry and ask, "Why did you change that, you said before you wanted X not Y?".  My answer is usually, "It LOOKS better now.", to which I get this glazed look of disbelief. They think that if I did it once a certain way before, then that's the only correct way of doing it in the future. To change it later on down the road means I don't know what I'm doing. A lot of programmers don't understand art, or the objective nature of design, to find both qualities in one person, programmer and artist, is truly a great thing.

Yes, yes and yes. Drag and drop seps (well, how you get the file in there is really NOT the point can and will do an awesome job. And what in the world is a 100% sep? A sep with NO issues? Doesn't that depend on how you print it?

I 100% believe that it is possible for a computer to sep an image just as well if not better than a human given the same art to start with.  What hasn't really been touched on much here, is the color correction and bumps made to the art by the professional separator BEFORE seps are made.  Therein lies a good amount of the talent in my opinion.

Yes, I believe (with some proof actually) that a computer can sep better. It's just math. I ALSO believe that it's important that the image is super cool BEFORE the sep is started. We can call it bumps, color correction, etc, but also vectorization of blurry, block jpgs is a part of that and in general, being artistic with the art! The point is, if you have an image that looks JUST the way you want it on your screen, this thing I made will do an awesome job making it look JUST LIKE THAT on a shirt. So any crappyness will be faithfully reproduced.

So, assuming, for a moment, that you are indeed happy with the image you slide in, then I believe the seps you get out will make you just as happy. But moreover, I believe that it can't really be done any better (provided I get all the adjustment curves perfect, which I will in time, with measurements). It'll start out as super and will slowly approach perfect as I make tiny adjustments to curves. I believe that even though high end pro seppers may take some variables into account when sepping (as opposed to when printing!), the raw math power behind my seps will still blow that "human advantage" out of the water.

Maybe I'll be proven wrong. I obviously doubt that, or I wouldn't be so bold .

...but for now if it comes down to choosing, I'd still choose the human pro seps for piece of mind...

Yes! I totally believe that as well! Actually I said that before. There's hundreds of thousands of screen printers and they want seps. A lot of them don't know how to do that, because, well, they do not have time to mess around with it forever. They also don't trust sep programs, because, well, they already wasted money on them and precious time doing reburns and retries. They will likely rather spend the money on a person, someone they can blame, at least, if it doesn't work. Otherwise they'd have to blame themselves if it's wrong, considering the 3rd party app is just a machine.

Seppers will therefore always have a job. However, I believe that a pro sepper would be wise to use my stuff at least as a base to work off of. I mean... there would really be no need to adjust anything, because if there were, I'd add that reason in and that would no longer be a reason to adjust anything. Does that make sense?

Any changes I add will be live to the world immediately. If there's an improvement, it's available to anyone. I'll be doing massive analysis on the prints produced by it, and any adjustments will be immediately used to help screen printers get sexier prints.

So, all in all, whatever knowledge can be pooled, ALL screen printers now can benefit from it, without even knowing how it works. There's so many that don't have time for this nonsense. Why would they care about WHY something gets adjusted and HOW MUCH? All they need is for it to be done, somehow, and done right. That's it. They need to move on quickly to the next paying job, while at the same time producing droves of happy customers. Sports teams and local businesses have small orders that don't warrant the large expense of a pro. Does that mean they should all have bland shirts? Well, maybe not anymore!

But my thing is also for the turbo pros, the ones with the awards. It's supposed to be near-perfect (as close as at least a human can get, but probably closer). That's the goal anyway, and I think I'm 90% there. Some curves from measurements is all I need. Oh, that and a LOT of system code. UI elements, settings storage, smooth workflow, reports, etc, etc.

Btw, what has not been clear from ANY of my previous posts I believe is that it will also do flat seps (no gradients). This is for picking simple logos apart. It's awesome at that as well, but to turn it on (now) means changing the source... no checkbox yet .

[DDSol]

I removed the white base for now. It's pretty good, but there's certain things I don't like about it. One of those is that it's horribly slow on one of my systems (which is 2 years old). I have a plan to make it very fast (almost 0 time), but it's work and I can't do it right now. I just don't want to cripple the test for people who also have slower systems until I fix it.

[DDSol]

you have deleted or edited my posts again... only a child would do that.

DUDE, I DID NO SUCH THING!!!

NOW WHO'S THE CHILD!?!?!