TheKitchenAbode

13 hours ago, paulchoate said:

Great advice BUT I don't want to turn off layers. I want to use 3D plants. I want my framing on. So, just tell me what I need to do to get it to run faster! PLEASE!

That's a tall order!!! What are the hardware specs of your current system? Maybe there is something that can be done but if your system is already reasonably configured then it's unlikely you will be able to reduce the lag enough. Again need to know your current specs.

Not sure why you are so adverse to turning of layer sets, this is the most effective way to improve performance and it does not cost any money to do so. It just involves a bit of front end effort to set-up/define the different working sets and then save this as your template. Like you, I have all kinds of 3D plants and furniture in my plans but if I'm working on roof planes there is no purpose in having those items displayed, they only clutter-up the view and slow things down and they can also be distracting.

It's important to understand that there is no software that can handle an infinite amount of complexity, they all slow down at some point in time. Breaking things down into more manageable chunks is the only effective way to deal with this and this technique is used in every industry software or otherwise.

Just now, PerryPlansPro said:

Where do you suppose I start in order to figure out the best solution?

To assist you need to describe precisely the issue you have and what you are attempting to do when things slow down. Better still is to post a problematic plan.

Looking at your system specs I'm not to sure there is much room for improvement as far as hardware goes. Your RTX 2080 can't be the problem and you have 32 GB of ram, both of these are far more than what CA needs. Your I7 7700 is a fairly good processor so upgrading this may not provide any significant performance boost, maybe 20% on average, also to upgrade this you will need a new motherboard that will add to the cost.

As others are suggesting, managing the display layer options to reduce display complexity is an easy one, especially in plan view if 2D panning and zooming lags. This can also help with 3D camera views but when choosing the things to turn off you need a different approach. In 3D things like textures, the number of active lights, the sun setting and shadow settings have a significant impact, if these are not needed then turn them all off.

In 2D plan views everything displayed are vector graphics and to rescale these each time you move around is highly CPU intensive. This is one of the major problems with say 3D plants, their cad blocks are way too complex, all those vector line details have to be recalculated when scaling. This is similar when additional content is imported such as high res pics or PDF content, it's not that the files themselves are a problem it's what has to be done to rescale them each time you pan and zoom.

You can also turn-off as many auto-rebuilds as possible so this is not happening to the entire model every time you make a change.

1 minute ago, Joe_Carrick said:

I think its extremely import to recognize the following:

• The GPU handles almost all of the 3D View processing for "PBR", "Standard" and "Vector" Renders.  The speed of this is dependent on the face counts, GPU speed, etc.
• The CPU handles all the rest of the graphics.  IOW, 2D CAD (Lines, Arcs, 2D Blocks, Text, Rich Text and Macros).  Optimization of 2D CAD can be done by limiting what's processed during Pan/Zoom as well as what Hardware is available (CPU, RAM, etc.) and how the software utilizes that hardware. 

IMO, any discussion of performance needs to be relative to 2D vs 3D - they are not the same.

Excellent point. When discussing this it is really important to differentiate between 2D and 3D performance as they involve distinctly different processes.

It's also important to recognize that even in 3D CA has to do a lot of computations when one makes changes to their model and these changes are primarily CPU based and they need to be completed before the graphical portion can be updated. 

13 hours ago, PerryPlansPro said:

I've taken a lot of different advice on this issue and nothing seems to make a significant difference.

This is most likely due to the fact that much of the given advice is often based upon broad generalities concerning computer performance. Though valid it does not reveal the true picture of what's going on in respect to a specific software package, unfortunately this can lead one to misinterpret the role a specific piece of hardware has and its impact on performance. For example, if your monitor is 60Hz and you find the movement on your monitor to be jittery then there may be nothing gained by choosing a graphics card solely on the basis that the new one can put out 120 Fps. First, a 60Hz monitor can only display 60 Fps and no more, secondly all graphics generation involves the CPU and thirdly how the software is written is all part of the involved process. To make the appropriate choice one needs to understand what role each of these have, is it an inherent limitation in the software programing, is the CPU to slow and can't feed the GPU fast enough or is the GPU not fast enough. For CA it can be even more complicated as there are many computational processes that must be done before anything is done on the graphical side, so is it the this that's really the issue, maybe it has nothing to do with the graphical processing/hardware.

1 minute ago, AgChief said:

 

Latency/unresponsiveness has always been a huge aggravation for me with CA.  This simple test took my system about 6 seconds each to increase and decrease magnification.  How can a simple single (albeit large) Psolid bog down the software like this?  Is this typical with everyone's system?

Everyone has the same(similar) experience. It bogs down the system due to the fill, it's set at 1/16" which generates a huge number of vector lines that must be regenerated each time you pan or zoom. This represents what happens as a plans complexity increases, each wall, framing member, cad object and symbol are in plan view vector lines. It was also designed to demonstrate that the number of surfaces are not the only contributor to lag, as this object only has 12 surfaces.  

Also, if when the camera is open and you have changed the display layer make sure to save the camera before closing it, otherwise it will use the prior one.

32 minutes ago, jorgearaya said:

I’m not an expert, but I use several video cards.

Nvidia Quadro is a professional grade card, I run that on a Lenovo P52s and Chief works fine, I see issues with AutoCAD it doesn’t like it.

Radeon Pro 555 also a professional grade, and runs all just fine no lags (Chief, ArchCAD, AutoCAD)

Basic intel integrated card seems to do great too!

Chief recommends “gaming” cards not professional grade. Nvidia gtx 1080 or AMD’s

Everything always runs fine as long as the plan remains below a certain level of complexity. Would be interesting if you would run this P-Solid stress test with your graphics cards to see how they perform.

Once the plan is loaded, first select the full screen icon and then record the time to reduce it by one magnification. Then reset it back to full screen and then record the time to increase it by one magnification.

P Solid Stress.plan

Just now, Michael_Gia said:

Would you say that as a general rule, turn all lights off, and only mess with the emmisivity of the materials to illuminate an interior scene? And same for an exterior scene?

Your PBR skills are off the wall by the way....

Thanks Michael. Having the lights on or off depends on how you wish the scene to look. I just wanted to demonstrate that X11 PBR can generate a half decent scene with just the sun only and a few minor material adjustments. The big advantage in not having lots of lights is that the PBR runs real fast and with just the sun it simplifies your lighting adjustments as there is only the sun to adjust.

Same for the exterior, the one I posted yesterday(a few posts above) only used the generic sun, no other lights. The model has 2.5 million surfaces. 116 3D trees/plants, with just 1 light source everything runs very fast.

I know many users are struggling with CA's PBR, but it's not really that difficult. It just takes a bit of controlled experimentation to develop an understanding as how the sun and lights work on their own and when together. Materials are only adjusted once the lighting is correct, not before.

5 hours ago, HumbleChief said:

Nice PBR's as usual Graham but curious about the reference to only the generic sun? Do the other ceiling lights in the shot not account for any computing needs...or??

The other lights are not turned on, the only light source in the scene is the generic sun. The recessed lights look like they are on because their lens material has emissive set in the material properties. Only lights that are actually turned on require computation.

Here's an example of an interior PBR that only uses the generic sun, no other lighting.

When using just the sun the interior ambient is going to be dictated by the amount of glazing, the more glazing the more ambient light. You can increase the ambient by cranking up the suns lumens but this will also crank up the direct light, find the right direct light level and If the interior ambient still appears low just add a small amount of emissive to the ceiling and wall surfaces. In this scene the ceiling emissive is 0.13 and the walls are 0.20.

As only the sun needs to be computed this PBR runs real fast, about 5 seconds to regenerate after a change, also rotates, zooms and pans smoothly. This model has 2.4 million surfaces and contains 116 3D trees/plant, it runs smooth and glitch free.

Just for interest, I took an X11 PBR, made a few basic adjustments to saturation and light levels in Photoshop. Here is a direct comparison to a scene that was generated in another rendering program.

The exterior X11 PBR only uses the generic sun, no other lighting.

Though the other scene is of better composure, most of the lighting qualities between the two are comparable.

It gets a bit more complicated if there is a material region spanning more than one floor. The material region even though it spans more than one floor is assigned 0nly to the floor based on it's start point when first drawn. Now if you try to paste a window from floor one onto floor two where the material region is it causes a problem as it seems to take the material regions floor assignment to locate the window, so it places it on floor 1 instead of floor 2.

13 minutes ago, Doug_N said:

The material region behaviour is related to how windows work in X11.  For the first time, you can move a copy of a window installed on the first floor to the second floor.  The window itself, however, is still a first-floor window even if now installed on the second floor.  If you do a window schedule, you will see that the window is listed as a first-floor window.  What is more, if you use a material dropper on a window that was actually installed on the second floor to move object properties to the first floor copied window, the window will drop to the first floor leaving the window casing, and the opening in the wall in place.  The "moved" target window will be embedded in the first-floor wall with no frame and no opening.  

 

 

So if you create a material region over the second floor that covers both of the natively installed window and the copied window, in some cases, only the natively installed window will be cut in the material region.  This behaviour does not seem to be consistent, so what is going on is somewhat obscured.  

Just played with this, good catch. It only seems to happen if you do a copy paste in place and then drag the window up onto another floor. If you do a copy and then paste it on another floor then everything seems to work out fine.

If it was working ok with X11 the other day then it might be something in that plan that is causing the problem. Try another plan to see if it happens again. You could also post that plan so we could see if we also get that message.

Library Analysis - I have just finished an extensive assessment of the processes involved during library use.

When the library is being accessed there are 3 processes involved, the first being accessing the disk drive to retrieve the symbol/texture, the second process involves the CPU to process the symbol/texture and the third process involves the GPU to display the model/texture.

Disk Drive Accessing - disk access times, regardless of symbol/texture complexity it was always under 1 millisecond. Also, under most circumstances once a symbol/texture had been accessed for the first time it remains in memory so subsequent accessing of the symbol does not require accessing the disk drive.

CPU Processing - this is where the bulk of the time is spent. Essentially the CPU processes are no different than those involved when generating a camera view. CA must generate the images of the symbol/texture for display in the selection panel and the preview panel, which are basically mini camera views. As symbol/texture complexity increased the CPU processing time increased accordingly.

GPU Processing - Though involved as it would be in any camera view the time to process was relatively minor. The GPU really only comes into play once the model has been generated/displayed and it then takes over when the symbol/model is being rotated or zoomed in on, no different than what happens in a standard camera view.

As revealed in prior stress tests, the CPU plays the most significant role in overall processing time and it's importance becomes even greater as complexity increases.
 

Just played with this. It does not seem to be related to an Attic wall. It happens on any wall if the window intrudes into the floor or ceiling platform. It seems that the material region does not recognize objects in those areas.

Your symbol does not work correctly because for some reason, the actual symbol size and the bounding box size/position is mismatched, maybe when it was first converted to a CA symbol something got corrupted.

Suggest just placing your symbol in a plan, not in a cabinet, adjust the symbols size to be proper, ignore the bounding box. Then convert it to a symbol, then place it in the plan again and open up the symbol and set the origin offsets and bounding box as they should be to insert into the cabinet door opening size. Save this to your library and all should be ok.

3 minutes ago, YoderW said:

but that should be stretching to fit anyway and doesn't fix it for my original symbol. 

I believe you will find that when a symbol is first inserted into a cabinet that CA does not size it to fit. It checks the symbols defined size and if the cabinet opening is large enough it will insert it at that size, if one of the dimensions is to great it pops up a message that the cabinet opening is not large enough and to resize the cabinet. The symbol stretch zones only come into play after the symbol has been inserted and you change the cabinets size.

That will be very tricky to do. CA's standard text functions do not support 3D. They do have one 3D text set available in the bonus catalogs but it just seems to be some generic block type text.

Not sure but maybe sketch-up might allow you to design the text portion and then you could import it into CA as a symbol.

14 hours ago, Chopsaw said:

Graham,  This has been happening to me as long as I can remember.  It does not seem to matter where I have my PDF's printed.  You must have your printer calibrated to Chief's PDF output somehow which must be different from everyone else's general default settings.  Unfortunately that is a big investment for someone just starting out.

I don't use any special calibration hardware/software, just do this by eye using some pics that I'm very familiar with and some free test patterns. What's important is that your monitors brightness and colour settings are independent of the actual pic's data, so the pic can look great on your screen but is in fact too dark. When you make changes in brightness, color, etc. in say Photoshop that data is stored in the pic and will be interpreted properly by the receiving device such as a printer. There may be something else going on but the first place I would take a look at is if your monitors settings such as brightness, gamma, color saturation is reasonably balanced. This can be done directly in your monitor settings or through your graphics card control DBX. When you do this you must face the fact that the proper settings may not always result in your screen looking right according to your personal preferences, many prefer overly saturated colors and higher than normal contrast, which is fine but when working with pics this will give you a false impression as to how your pic will really look when reproduced according to it's true data profile.

your posted PDf's printed out fine for me. Is this something that has just started to happen or does everything outputted from your computer print overly dark?

An example would be great. Maybe it's the printer settings that are the issue, most printers have setting controls that allow you to adjust brightness, contrast, colour saturation and more so you can align/calibrate the printer with your screen.

1 hour ago, jjifmeyer said:

Wish the search in the 3d library was a little better.

The library search could be better but in your situation I think you were misinterpreting the fact that when you search it only searches the libraries currently installed on your system, there is no online search feature that looks at all of the available library catalogs.

3 minutes ago, Kbird1 said:

Thought you might like it 

 

It's one of the few plans I have seen that causes all 6 cores on my 5820k to hit 100% ( using CoreTemp 1.12.1) ....going to a 3D overview

 

Once in 3D it is no issue though,  but my GPU hits full boost and stays there while panning around , most CA 3D isn't that intensive.

Those 6 cores are maxing out all the time, it's just that due to lower complexity the duration is very short so they may only be maxed out for a few milliseconds.

Same for my HD620, maxes out when rotating and zooming in 3D. 

20 minutes ago, Kbird1 said:

 

Hmmmm that's curious , what kind of error? I don't get any....

 

The 3'x5' Block is actually from a Plan someone posted on the Forum a year or two ago , I think they got it from 3D warehouse IIRC and while it does look good in 3D , I think the amount of CAD lines CA has to draw on a Whole Lawn might be what is killing it......pretty much the same as fill a whole Terrain Perimeter with a hatch fill spaced at 1/8".

 

M.

Tried again and it loaded in.

That one is more brutal than my Stress Test P-Solid as it adds a crazy number of surfaces, around 16 million. Now plan panning and zooming are impossible but opening up a camera view takes considerable time.