'X Position' fade jumpy

Hi there,
Wondering if theres a quick fix here...

I'm trying to slide video across stage to match a scenic automation but X position seems to be very low resolution and this results in a jerky motion.
The movement is about 1/2 of the 1024x768 screen width, using an 8 sec timed fade from the Hog III.

It looks like it jitters back and forth as it goes along, is there a way to increase the resolution of the move?

i talked with flying pig and they think it is the catalyst, this would tally with the manual movement resolution of X position when programming so it looks like they might be right...

Any ideas i might try?

Thanks,

Sam Hopkins
Projection Engineer & Programmer
Sunday In The Park With George
Studio 54 new york

Try using the "Movement Speed" channel....this is exactly the kind of issue it is designed for.

On H3 it is called "Image Speed" under BEAM.

***EDIT- Not to be confused with "Play Speed" which is something else entirely***

Hope this helps :)

Quote:

---

Originally Posted by Samlight

i talked with flying pig and they think it is the catalyst

---

who told you that ?

It's not the Catalyst it's the DMX protocol it self not the catalyst .

try record a preset in catalyst HUD for start of X position

then trigger the the next Catalyst preset using DMX Preset Trigger .

Funnily enough the 'Head of console development' told me that when I rang to ask if it was his dodgy fades... I can't see how its 'the DMX protocol itself' though, DMX is more than capable of the resolution I'm looking for.

I'll try movement speed today, thanks.

Sam

Actually, I know a number of people who have run into this problem on the Hog 3, it's not the greatest of DMX processing. I know they have improved it recently, but I have seen quite a few people pulling their hair out over this same issue on a few shows. Not just with media servers either.

I always get blamed for this - and its not me.
its not dmx itself - its not catalyst - ( unless you are overloading catalyst )


its the update rate of dmx from the console - or their inconsistent fade calculations.
Its also the fact that dmx is not synchronised to video - and they never output dmx fast enough to keep up with 30fps or 25 fps video.

Several years ago - the hog3 also used to be only 12bit on 16 bit channels.
They may have played with these algorithms.

I have added graph plotting to the software recently to PROVE it isnt me. see image below - you can plot a graph of incoming dmx.
Because its also a problem with fading intensity.
I got blamed for poor fading on several shows with hog2- but the hog2 only outputted dmx at 20fps...

I also added movement smoothness overrides on each layer- so that if the movement from the console isnt good enough - and you cant reprogram a show - you can add movement smoothness anyway.

You can also enable movement smoothness on the keystone and aspect ratio channels. On a show in london last year - they just couldnt get smooth enough dmx from a hog2 to make this seemless-

--

At the bottom of each layer in the HUD - there is a number with a pulldown menu-
image shown below-

You can enable movement smoothness overrides on intensity or movement channels, with different amounts of smoothness.

---


Richard

Thanks Richard for your response.

I tried the things mentioned, looks like movement speed it will be.

The movement smoothness overrides don't seem to work with a 0 count fade, or rather, it shouldn't work on an instant cue but it does! This gives me issues, so it has to be something I can engage from the console only when required.

Movement speed appears to help.

Thanks again, happy to report that other than this issue my 3 cats are working wonderfully as usual.

Cheers,

Sam

Oh, PS, can anyone recommend a reasonable USB video or vga input device to use? The director has requested we try a trick involving input from a PC screen... don't want to blow the bank just checking out whether it works, but happy to spend a couple of hundred bucks to get a feed in and try it... any cheap and cheerful favourites?

There are some Firewire devices out there that are essentially video capture interfaces.

I have had good luck with both the DataVideo DAC-100 for composite/s-video inouts and the Canopus TwinPact-100boxes as well for this plus RGB + VGA inputs.

Link to DataVideo:
http://www.datavideo.us/products/dac_100_main_page.htm

Link to Canopus:
http://www.canopus.com/products/Twin...winPact100.php

Quote:

---

Originally Posted by Marty Postma

Link to Canopus:
http://www.canopus.com/products/Twin...winPact100.php

---

thanks marty.

this looks interesting to digitise vga- ill get one to check.

havent seen this before.

Quote:

---

Originally Posted by samsc

thanks marty.

this looks interesting to digitise vga- ill get one to check.

havent seen this before.

---

No problem...Mainlight uses them all the time on the VLSE computers for SoftLED...very reliable.

The Hog 3 issue was something I raised with Robbie Bruce at High End some time ago. I understood this issue has been resolved..

Hugh

Quote:

---

Originally Posted by Spam Butterfly

The Hog 3 issue was something I raised with Robbie Bruce at High End some time ago. I understood this issue has been resolved..

---

This was my understanding as well....I think this was fixed a while back

Now thats interesting... it was Robbie I rang to check with.

S

Quote:

---

Originally Posted by samsc

thanks marty.

this looks interesting to digitise vga- ill get one to check.

havent seen this before.

---

Hi
I will need to put on the video input layer a vga output (1024x768 I guess...) from a 8 hours long Flash-based "interactive" animation running fullscreen on a laptop during a week-long show. This could be a very good solution.
Have you tested?
Works better than other scan conv.?

Thanks
Stefano
Raggi di Luce

MacPro Quad
Catalyst m165
Ati x1900

Quote:

---

I will need to put on the video input layer a vga output (1024x768 I guess...)

---

you have to test how this affects what you need it to do.

its advantage is that it is a firewire convertor- converts to dv - so works with laptops.

I think we've all had the phone call where we're told "It's not us, you just don't do it right." GOD! That pisses me off.

Running a cue with fade timing will directly triggers the DP2000 to complete a fade. Using a fader however requires the console to transfer new state data which must be processed by the DP2000. As Richard mentioned, the fade characteristics of the faders are a bit different from what a user might expect and can sometimes produce choppy results. This is actually inherent in many control surfaces. We just don't tend to notice it when it's only with light. The positioning of video is much more complex.

If you look at the Hog 3 forum, you'll see for yourself a number of posts relating to this.

I have dealt with this tons using automated lights to track performers. The only solution is to mark out a performer path and create something like 10 steps. Each step is a cue with a tenth of the movement. I then record the cues with "damped" fading. Since the performer never really stops, the damped is good because the beam will slow down subtly as it approaches the end of a step. The good thing though is that if you're at cue 1 and the performer walks too damned fast, you just step through to stay a cue ahead.

The way this works is that instead of fading with each individual movement, a GO which jumps several cues will cause the DMX processor to calculate the "current" state at the GO press and then render the final state before calculating a DMX transition rate.

It gets a bit simpler on Catalyst though since there is an image speed. All of us from the old days when the only way to set position speed was to dial in a speed channel know this already. Here's a solution that works.

On the hog 3, set the X pos timing to zero in the x-fade and set your image speed to whatever it needs to be to accommodate each step of the whole move Also set the timing to zero. Then in each of the 10 steps progress the X pos a tenth. The image speed will track through and the DMX signal for each step to Catalyst will be instantaneous.

Therefore, the product with higher resolution (Catalyst) will transition smoothly and you won't have to worry about the dynamic unreliability of DMX.