Revit 2011 & Autocad 2011

[TMA_80]

Hi!

Ok, i know, I'm posting too much "other products" links....at least we keep updated ( Autodesk is even launching 2011 products !?)

So an update to the last year post 😉 :

Revit 2011
http://autodesk-revit.blogspot.com/2010/03/whats-new-in-revt-architecture-2011.html

this version seems to fix the ui interface problem, complete the "missing" options for the new freeform modeling tool (which i think is great) ...(no multithreading yet ?! )..and textured view option ...etc.

Autocad2011
http://www.cadalyst.com/cad/autocad/autocad-2011-a-look-what039s-new-13200

[Anonymous]

Mats_Knutsson wrote:

Andri wrote:
I have use both revit and archicad. And I find that revit develop their render engine. In Revit 2008 they use accurender, but in revit 2009 they start to use mental ray. This make a gap between archicad and revit. Different from revit, archicad doesn't develop their render engine. I think if archicad use mental ray ot would be great. Because many people in my country proud of revit because its render.

Apart from the obvoius fact that ArchiCAD soon will have a new rendering engine I look at the huge base of single person or small companies that can't affort having their bulk designing tool render for half the day. There is a reason ArtLantis is popular. IF....a rendering engine inside the cad-program would be extremely fast...then OK...

I think both Revit and Archicad has their own strength..I am also have tried both Revit and Archicad well, but here I want to give you to know the strength of Parametric Change Engine in Revit that works live and full bi-directional associativity..
There is some amount of confusion regarding full bi-directional associativity and it may be useful to spell out what it means. I don't really want to focus on the "my software is better than yours" aspect but would rather describe the meaning of terms "live" and "bi-directional associativity". Any product making claims to support bi-directional associativity has to be able to support examples of parametric and associative changes listed below.

Many existing products have some ability to update elements of a design when other elements change. However with an exception of Revit in many cases such updates are not automatic. Every case when a user has to take an explicit action to do an update creates a possibility of error and uncoordinated design documents. It also necessitates additional work by an end user. We also should not confuse the ability to display or edit a single underlying data model in multiple views with full parametric associativity between various elements of design.

There are 3 main classes of elements in any building design 
(A) Building components (walls, roofs, doors, windows, floors, etc.)
(B) Views including schedules and sheets
(C) Annotations (text notes, dimensions, spot elevations, etc.)

I do not want to diminish benefits of other products, they are quite good at what they do. 
On the other hand Revit is the only product on the market today which was engineered from the ground up to provide full bi-directional associativity between all 3 main classes of elements.

Below are the examples of this associativity and corresponding parametric change propagation from elements of class to another. 

Building components to building components
a. Move one wall and connected adjacent walls adjust to become longer or shorter
b. Move walls and a floor adjusts to cover area enclosed by walls
c. Raise/lower a roof and attached walls grow or shrink
d. Thicken a wall and door frames adjust to new thickness
e. Raise a level and all elements placed of this level will follow.

Building components to views
a. Change to a building component is automatically reflected in all graphical views without additional user actions
b. Move walls and room schedule updates room areas
c. Add or remove building components or change parameters of existing elements and schedules update automatically

Building components to annotations
a. Change geometry and dimension value updates
b. Move things higher or lower and spot elevation reflects new heights
c. Move walls and room tags update displayed area values
c. Changes to properties of wall, windows, doors, etc. are automatically reflected in their tags

Views to building components
a. Any graphical view (plan, elevation, section, callout) may be used to effect a change to building component
b. Changes to building components may be made by editing their parameters in schedules
c. Changes to view phase or level of detail automatically reflected in display of all building components shown by this view

Views to other views
a. Move section or detail view backward or forward and callouts move with their parent section 
b. View and drawing schedules (view/drawing lists) may be used to change properties of other views and drawing

Views to annotations
a. Change view scale and all dimensions, text notes, etc. adjust to maintain their sizes on printed output
b. Place a view on a drawing sheet and view tags update to reflect sheet number
c. Change view scale and scale tag in view title on sheet updates

Annotations to building components
a. Change dimension value and building component changes accordingly
b. Changes to property values shown by tags automatically propagate to building components
c. Change elevation value displayed by level tag and level moves up or down
d. Impose dimension equality constraint or lock dimension value and building components behave accordingly

Annotations to views
a. Flip direction of section view tag and view forward direction flips.

Annotations to annotations
a. Change sheet number in a titleblock and the change will propagate through drawing to views placed on this drawing and then to their view tags (section and callout heads). 

All these examples are made possible in Revit not only because it has a patent pending Parametric Change Engine (PCE) in the middle of its software architecture but also because all Revit's elements are implemented with parametric change in mind. There are countless other examples made possible by the PCE and the unifying notion of associativity between all 3 kinds of design elements. 

Unless a product is implemented from the ground up with a PCE type of architecture its implementation may exhibit some examples of associativity but its change propagation capabilities are bound to be limited.

[Anonymous]

Rikhardus wrote:

Mats_Knutsson wrote:

Andri wrote:
Unless a product is implemented from the ground up with a PCE type of architecture its implementation may exhibit some examples of associativity but its change propagation capabilities are bound to be limited.

Agreed. I really could appreciate how in Revit you can have a number of plans sections, elev's etc open on the screen. Pick an element in any view and it highlights and moves etc. in all other views as well. I also think the fact that you can move a wall in the floor plan and know that other elements on other plans (Ceiling grid etc) change with out having to turn on all layers to stretch them with the marquee tool.

ArchiCad has a smoother 3d view including textures (if it only had some additional modeling tools). Many of my clients appreciate online meetings and so forth viewing the design in a dynamic way in the 3d window. I don't know but Revit might be getting better in the 3d views.

[owen]

Rikhardus wrote:
All these examples are made possible in Revit not only because it has a patent pending Parametric Change Engine (PCE) in the middle of its software architecture but also because all Revit's elements are implemented with parametric change in mind. There are countless other examples made possible by the PCE and the unifying notion of associativity between all 3 kinds of design elements. 

Unless a product is implemented from the ground up with a PCE type of architecture its implementation may exhibit some examples of associativity but its change propagation capabilities are bound to be limited.

The features you have listed are not exclusive to Revit .. many of them are possible in ArchiCAD - but not all.

Historically I think Graphisoft have not pursued a fully-associative model approach due to the performance issues this would entail - the software and hardware have just not been up to the task, so they have maintained a responsive application at the expense of true parametrics throughout the application.

It made sense at the time .. users constantly frustrated at the performance of the program they are using day after day is not a good thing. But I think the hardware capabilities are now rapidly outpacing what the software is capable of throwing at it - I do not know if the ArchiCAD engine is capable of incorporating the associativity to the level of Revit in the future, it really depends on how the program is structured (i.e could things like associative walls be easily enabled or not)

I think this is the area where Revit will pull rapidly away from ArchiCAD. If performance is not as much of an issue, then i think the Revit approach to model associativity is better in the long run.

[Anonymous]

We should have a really robust 3d modeling and object creation environment if we will not be able to compete in the long run on the change management aspect. One with the power, ease and flexibility of Bonzai 3d with improved documentation characteristics.

[Anonymous]

Building components to building components
a. Move one wall and connected adjacent walls adjust to become longer or shorter
b. Move walls and a floor adjusts to cover area enclosed by walls
c. Raise/lower a roof and attached walls grow or shrink
d. Thicken a wall and door frames adjust to new thickness
e. Raise a level and all elements placed of this level will follow.

Except that if you build your revit model this way, it quickly becomes over constrained any besides crawling to a standstill on even a liquid cooled overclocked gaming machine, becomes difficult to unconstrain so you can actually move a wall. Good if you start out knowing what the design is, bad if you want to actually design.

Building components to views
a. Change to a building component is automatically reflected in all graphical views without additional user actions
b. Move walls and room schedule updates room areas
c. Add or remove building components or change parameters of existing elements and schedules update automatically

Revit is only one way on schedules. Archicad is bidirectional. Archicad clearly wins on this one. Doors and windows can also have tags that update on elevations. On revit, this must be done manually for each view.

Building components to annotations
a. Change geometry and dimension value updates
b. Move things higher or lower and spot elevation reflects new heights
c. Move walls and room tags update displayed area values
c. Changes to properties of wall, windows, doors, etc. are automatically reflected in their tags

Duh.

Views to building components
a. Any graphical view (plan, elevation, section, callout) may be used to effect a change to building component
b. Changes to building components may be made by editing their parameters in schedules
c. Changes to view phase or level of detail automatically reflected in display of all building components shown by this view

Views to other views
a. Move section or detail view backward or forward and callouts move with their parent section
b. View and drawing schedules (view/drawing lists) may be used to change properties of other views and drawing

Revit only allows one reference symbol to point to one view. Archicad allows a detail to be referenced any number of times.

Views to annotations
a. Change view scale and all dimensions, text notes, etc. adjust to maintain their sizes on printed output
b. Place a view on a drawing sheet and view tags update to reflect sheet number
c. Change view scale and scale tag in view title on sheet updates

Both programs do equally well. Archicad allows text to be non scale specific, useful for site plans.

Annotations to building components
a. Change dimension value and building component changes accordingly
b. Changes to property values shown by tags automatically propagate to building components
c. Change elevation value displayed by level tag and level moves up or down
d. Impose dimension equality constraint or lock dimension value and building components behave accordingly

Annotations to views
a. Flip direction of section view tag and view forward direction flips.

Annotations to annotations
a. Change sheet number in a titleblock and the change will propagate through drawing to views placed on this drawing and then to their view tags (section and callout heads).

All these examples are made possible in Revit not only because it has a patent pending Parametric Change Engine (PCE) in the middle of its software architecture but also because all Revit's elements are implemented with parametric change in mind. There are countless other examples made possible by the PCE and the unifying notion of associativity between all 3 kinds of design elements.

This is total BS. Revit may patent their version of a parametric engine, but it is neither the first, nor the best. CATIA, pro-engineer and Archicad have been around longer and are more robust and functional.


Unless a product is implemented from the ground up with a PCE type of architecture its implementation may exhibit some examples of associativity but its change propagation capabilities are bound to be limited.


same point above....also add that the parametric abilities of Revit do nothing for the clunky interface. In fact the parametric bend on Revit is what makes it hard to design with, which is why every Revit office has Sketch-up for designers. Not so with archicad.

Philosohically, Revit assumes design is a linear process which consist of diminishing granularity and that you just go around making relationships and constraining them. It does not do fudge very well which means you have to know what you are doing before you do it.

ohh. and it crahses all the time unlike archicad, which in my mind is another big problem.

[Anonymous]

Tomas wrote:
Except that if you build your revit model this way, it quickly becomes over constrained any besides crawling to a standstill on even a liquid cooled overclocked gaming machine, becomes difficult to unconstrain so you can actually move a wall. Good if you start out knowing what the design is, bad if you want to actually design.

Not really. All those constrains are usually default ones (OOTB).
Overconstrains are when users lock mulltiple components manually.

Tomas wrote:
Revit is only one way on schedules. Archicad is bidirectional. Archicad clearly wins on this one. Doors and windows can also have tags that update on elevations. On revit, this must be done manually for each view.

Wrong again.
Updates in schedules is relected in the model. The opposite is also true.
Propagation of parameters is automatic in all views (schedules included).
No update command required.

Tomas wrote:
Revit only allows one reference symbol to point to one view. Archicad allows a detail to be referenced any number of times.

In RAC, a detail can be referenced in any number of times.
A live view cannot. But then why should it?

Tomas wrote:
Both programs do equally well. Archicad allows text to be non scale specific, useful for site plans.

There is a model text component that is not scale specific.
And yes, revit has very basic text formatting, compared with AC.

Tomas wrote:
This is total BS. Revit may patent their version of a parametric engine, but it is neither the first, nor the best. CATIA, pro-engineer and Archicad have been around longer and are more robust and functional.

Revit is a bastard son of pro-engineer. It's modelling engine was based on very basic, yet powerfull component relashionships.
True it was not the first, but it got a fresh clean start.
IMHO, the beauty of it was the lack of CAD legacy issues.

Tomas wrote:
same point above....also add that the parametric abilities of Revit do nothing for the clunky interface. In fact the parametric bend on Revit is what makes it hard to design with, which is why every Revit office has Sketch-up for designers. Not so with archicad.

That is all about the end-user IT literacy.

Tomas wrote:
Philosohically, Revit assumes design is a linear process which consist of diminishing granularity and that you just go around making relationships and constraining them. It does not do fudge very well which means you have to know what you are doing before you do it.

Well, I can create a loose cabinet family, or I can create one that is attached to a wall, making it's
placement and control more obvious. All options are valid, but one is more obvious. That is going around making relationships and constraining them?
As one new revit user (came from autocad) once asked: "does this means that what I model in revit is actually what means to be constructed?".
That basic workflow principle in revit is that the model should reflect the final construction. I can see no harm in that.

Tomas wrote:
ohh. and it crahses all the time unlike archicad, which in my mind is another big problem.

Not that frequently.

Tomas, this is just a correction of your revit analysis, not a comparison with AC, of which I know very little.

[Anonymous]

Revit is a bastard son of pro-engineer. It's modelling engine was based on very basic, yet powerfull component relashionships.
True it was not the first, but it got a fresh clean start.
IMHO, the beauty of it was the lack of CAD legacy issues.

The idea of a "sketch" window to edit profiles is a CAD legacy. I used to teach Pro-Engineer at RPI, and when i first saw Revit, I recognized the Pro-Engineer interface, down to the term pink lines for paths and shapes, sans the blue screen. No "fresh start" and the Pro-Engineer legacy issues, including the parametric engine, are exactly what limits the program.

In RAC, a detail can be referenced in any number of times.
A live view cannot. But then why should it?

Typical elevations, as in Pantry, elevator lobby, etc, typical details (cut live from model)

Tomas wrote:
Except that if you build your revit model this way, it quickly becomes over constrained any besides crawling to a standstill on even a liquid cooled overclocked gaming machine, becomes difficult to unconstrain so you can actually move a wall. Good if you start out knowing what the design is, bad if you want to actually design.

Not really. All those constrains are usually default ones (OOTB).
Overconstrains are when users lock mulltiple components manually.

OK, so everytime revit ask you to constrain the slab edge to the walls, that is the a user constraining. So you don't constrain, you can't move slab edge with wall. Being as there is no fence stretch, you have to move a wall then move a slab edge.
And over constraining is when multiple constraints cannot be resolved. Multiple components can have constraints and components can have multiple constraints, and the constraints can be resolvable. You just need to be a engineer.

Wrong again.
Updates in schedules is relected in the model. The opposite is also true.
Propagation of parameters is automatic in all views (schedules included).
No update command required.

I can't change the width of a door in a revit schedule without creating a family type or making sure all the doors have width as a type parameter. I cannot select a door in a schedule and go to it in the model either.

Well, I can create a loose cabinet family, or I can create one that is attached to a wall, making it's
placement and control more obvious. All options are valid, but one is more obvious. That is going around making relationships and constraining them?
As one new revit user (came from autocad) once asked: "does this means that what I model in revit is actually what means to be constructed?".
That basic workflow principle in revit is that the model should reflect the final construction. I can see no harm in that

Constraint based models are great if you are designing say, a jet engine. All the parts fit together and interconnect in a way that architectural systems don't. Tolerances in architecture is in the order of large fractions, ie, 1/2", and maybe 1/16" on millwork and are picked up by the layering of systems.
The purpose of the design of a jet is to achieve this interconnectedness, not so in architecture. Being able to move a wall without having to wait for the program to resolve all that is entailed to move the wall is part of design. You may not know you need a cabinet attached to a wall and you may need to move it off to the side (off the wall) while you figure a few things out. Revit assumes what is more obvious on too many things, which I guess is why some people like it, and why I hate it.

Not that frequently. (crashes)

Even CADlyst list this as one of the complaints.

Tomas, this is just a correction of your revit analysis, not a comparison with AC, of which I know very little.


Well, i disagree with your correction, obviously. I have used both, and I venture to say very few people who tout Revit as the next thing since swiss cheese have. It is better than AutoCad for sure, but it ain't all that. Dassault is about to debut it's BIM /Catia cousin. I venture to say that like ArchiCad it will also be better than Revit....but that will not change Autodesk marketing prowess which is equal to MacDonald's mastery of making the big mac, a palette-less combination of fat and starch, the most popular single food item on the planet.

[Chadwick]

I don't really want to 'pile on' so to speak and turn this into a Revit vs ArchiCAD thing yet again BUT I think a lot of the discrepancies that both sides have to do with someone being used to the workflow of another. Having used both, I can see the ups and downs of each. I understand why in Revit you would want to make a new family type for each door but unless you use Revit you would know that you could also make a door that has a width parameter that is instance based and therefore can be changed without having to create a whole new family. Now, I'll agree that the way Revit is set up that you wouldn't want to do this because type parameters are partly what drives the organization of families, elements, etc. I honestly have no problem with that workflow. I also really liked the workflow that ArchiCAD had back when I used in version 10. Also, I don't have any problems crashing in Revit and when I used ArchiCAD I crashed more often - but this could simply be getting lucky with the hardware in my computer. But I am rambling...

My larger point is some of these 'comparisons' are so subjective that they really carry little validity for the masses.

[Anonymous]

My larger point is some of these 'comparisons' are so subjective that they really carry little validity for the masses

A more objective comparison, ie number of mouse clicks would be equally meaningless because peoples perception on the strength of a platform is not perceived this way and the user may also be adjusting by altering the way they design to use less mouse clicks. Productivity would have an equally similar problem in that you would adjust how much time you spent getting things in a design the way you want them to suit a target profitability.
It is subjective, but that does not mean the comparison carries little validity. Often decisions are not made on the quality of a program but rather the perception on whether or not they are following some industry standard, or they don't bother to find out how the programs stack up. Then once they adopt the program they start drinking the cool aid.
Yes Revit has some good things and things it does better than AC, but it does a LOT of things poorly in comparison to AC.

and...in case you didn't notice, this is a Graphisoft forum, so what would be the point of a Revit Autocad thread that is not focused on a comparison?

[Anonymous]

i meant eg....not ie. ...that happened to be the method Microstation used to compare themselves to AutoCod.